code
stringlengths 87
55.2k
| code_codestyle
int64 0
349
| style_context
stringlengths 135
49.1k
| style_context_codestyle
int64 0
349
| label
int64 0
1
|
|---|---|---|---|---|
from __future__ import annotations
def lowerCAmelCase_ ( snake_case_ ):
create_state_space_tree(snake_case_,[],0,[0 for i in range(len(snake_case_ ) )] )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_,):
if index == len(snake_case_ ):
print(snake_case_ )
return
for i in range(len(snake_case_ ) ):
if not index_used[i]:
current_sequence.append(sequence[i] )
_A : List[Any] = True
create_state_space_tree(snake_case_,snake_case_,index + 1,snake_case_ )
current_sequence.pop()
_A : Union[str, Any] = False
_snake_case = [3, 1, 2, 4]
generate_all_permutations(sequence)
_snake_case = ["A", "B", "C"]
generate_all_permutations(sequence_a)
| 26 |
import argparse
import shutil
import time
from json import JSONDecodeError
from logging import getLogger
from pathlib import Path
from typing import Dict, List
import torch
from torch.utils.data import DataLoader
from tqdm import tqdm
from transformers import AutoModelForSeqaSeqLM, AutoTokenizer
from utils import (
SeqaSeqDataset,
calculate_bleu,
calculate_rouge,
chunks,
lmap,
load_json,
parse_numeric_n_bool_cl_kwargs,
save_json,
use_task_specific_params,
write_txt_file,
)
_snake_case = getLogger(__name__)
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ = 8,snake_case_ = 1024,snake_case_="val",snake_case_=None,snake_case_=False,snake_case_="summarization",snake_case_=None,snake_case_=1,snake_case_ = None,snake_case_="",**snake_case_,):
_A : Dict = str(snake_case_ )
assert local_rank is not None
torch.distributed.init_process_group(backend="""nccl""",rank=snake_case_ )
_A : Tuple = Path(snake_case_ )
_A : List[Any] = save_dir.joinpath(f'''rank_{local_rank}_output.json''' )
torch.cuda.set_device(snake_case_ )
_A : Union[str, Any] = AutoModelForSeqaSeqLM.from_pretrained(snake_case_ ).cuda()
if fpaa:
_A : Any = model.half()
# determine if we need to increase num_beams
use_task_specific_params(snake_case_,snake_case_ ) # update config with task specific params
_A : str = generate_kwargs.pop("""num_beams""",model.config.num_beams ) # AttributeError risk?
if num_return_sequences > num_beams:
_A : int = num_return_sequences
_A : Optional[Any] = AutoTokenizer.from_pretrained(snake_case_ )
logger.info(f'''Inferred tokenizer type: {tokenizer.__class__}''' ) # if this is wrong, check config.model_type.
if max_source_length is None:
_A : Optional[int] = tokenizer.model_max_length
if prefix is None:
_A : Tuple = prefix or getattr(model.config,"""prefix""","""""" ) or """"""
_A : Optional[int] = SeqaSeqDataset(
snake_case_,snake_case_,snake_case_,max_target_length=1024,type_path=snake_case_,n_obs=snake_case_,prefix=snake_case_,**snake_case_,)
# I set shuffle=True for a more accurate progress bar.
# If all the longest samples are first, the prog bar estimate is too high at the beginning.
_A : Optional[int] = ds.make_sortish_sampler(snake_case_,distributed=snake_case_,add_extra_examples=snake_case_,shuffle=snake_case_ )
_A : Dict = DataLoader(snake_case_,sampler=snake_case_,batch_size=snake_case_,collate_fn=ds.collate_fn )
_A : Optional[Any] = []
for batch in tqdm(snake_case_ ):
_A : Tuple = model.generate(
input_ids=batch["""input_ids"""].to(model.device ),attention_mask=batch["""attention_mask"""].to(model.device ),num_return_sequences=snake_case_,num_beams=snake_case_,**snake_case_,)
_A : Any = tokenizer.batch_decode(snake_case_,skip_special_tokens=snake_case_,clean_up_tokenization_spaces=snake_case_ )
_A : Dict = batch["""ids"""]
if num_return_sequences > 1:
_A : Any = chunks(snake_case_,snake_case_ ) # batch size chunks, each of size num_return_seq
for i, pred in enumerate(snake_case_ ):
results.append({"""pred""": pred, """id""": ids[i].item()} )
save_json(snake_case_,snake_case_ )
return results, sampler.num_replicas
def lowerCAmelCase_ ( ):
_A : Tuple = argparse.ArgumentParser(
epilog="""Unspecified args like --num_beams=2 --decoder_start_token_id=4 are passed to model.generate""" )
parser.add_argument("""--data_dir""",type=snake_case_,help="""like cnn_dm/test.source""" )
parser.add_argument(
"""--model_name""",type=snake_case_,help="""like facebook/bart-large-cnn,t5-base, etc.""",default="""sshleifer/distilbart-xsum-12-3""",)
parser.add_argument("""--save_dir""",type=snake_case_,help="""where to save""",default="""tmp_gen""" )
parser.add_argument("""--max_source_length""",type=snake_case_,default=snake_case_ )
parser.add_argument(
"""--type_path""",type=snake_case_,default="""test""",help="""which subset to evaluate typically train/val/test""" )
parser.add_argument("""--task""",type=snake_case_,default="""summarization""",help="""used for task_specific_params + metrics""" )
parser.add_argument("""--bs""",type=snake_case_,default=8,required=snake_case_,help="""batch size""" )
parser.add_argument(
"""--local_rank""",type=snake_case_,default=-1,required=snake_case_,help="""should be passed by distributed.launch""" )
parser.add_argument(
"""--n_obs""",type=snake_case_,default=snake_case_,required=snake_case_,help="""How many observations. Defaults to all.""" )
parser.add_argument(
"""--num_return_sequences""",type=snake_case_,default=1,required=snake_case_,help="""How many sequences to return""" )
parser.add_argument(
"""--sync_timeout""",type=snake_case_,default=600,required=snake_case_,help="""How long should master process wait for other processes to finish.""",)
parser.add_argument("""--src_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument("""--tgt_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument(
"""--prefix""",type=snake_case_,required=snake_case_,default=snake_case_,help="""will be added to the begininng of src examples""" )
parser.add_argument("""--fp16""",action="""store_true""" )
parser.add_argument("""--debug""",action="""store_true""" )
_A : Union[str, Any] = time.time()
_A , _A : List[str] = parser.parse_known_args()
_A : List[str] = parse_numeric_n_bool_cl_kwargs(snake_case_ )
if generate_kwargs and args.local_rank <= 0:
print(f'''parsed the following generate kwargs: {generate_kwargs}''' )
_A : Dict = Path(args.save_dir + """_tmp""" )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ ) # this handles locking.
_A : int = list(json_save_dir.glob("""rank_*.json""" ) )
if intermediate_files:
raise ValueError(f'''Found files at {json_save_dir} please move or remove them.''' )
# In theory, a node could finish and save before another node hits this. If this happens, we can address later.
_A : Any = {}
if args.src_lang is not None:
_A : int = args.src_lang
if args.tgt_lang is not None:
_A : Dict = args.tgt_lang
Path(args.save_dir ).mkdir(exist_ok=snake_case_ )
_A , _A : str = eval_data_dir(
args.data_dir,snake_case_,args.model_name,type_path=args.type_path,bs=args.bs,fpaa=args.fpaa,task=args.task,local_rank=args.local_rank,n_obs=args.n_obs,max_source_length=args.max_source_length,num_return_sequences=args.num_return_sequences,prefix=args.prefix,dataset_kwargs=snake_case_,**snake_case_,)
if args.local_rank <= 0:
_A : List[Any] = Path(args.save_dir )
save_dir.mkdir(exist_ok=snake_case_ )
_A : Tuple = gather_results_from_each_node(snake_case_,snake_case_,args.sync_timeout )
_A : Optional[int] = combine_partial_results(snake_case_ )
if args.num_return_sequences > 1:
_A : Optional[Any] = save_dir.joinpath("""pseudolabel_results.json""" )
print(f'''Saving aggregated results at {save_path}, intermediate in {json_save_dir}/''' )
save_json(snake_case_,snake_case_ )
return
_A : List[str] = Path(args.data_dir ).joinpath(args.type_path + """.target""" )
with open(snake_case_ ) as f:
_A : int = [x.rstrip() for x in f.readlines()][: len(snake_case_ )]
# Calculate metrics, save metrics, and save _generations.txt
_A : Dict = """translation""" in args.task
_A : Optional[Any] = calculate_bleu if calc_bleu else calculate_rouge
_A : Tuple = """bleu""" if calc_bleu else """rouge"""
_A : Dict = score_fn(snake_case_,snake_case_ )
_A : List[Any] = len(snake_case_ )
_A : Optional[int] = time.time() - start_time
_A : Dict = round(runtime / metrics["""n_obs"""],4 )
_A : Dict = num_replicas
# TODO(@stas00): add whatever metadata to metrics
_A : Any = save_dir.joinpath(f'''{args.type_path}_{metric_name}.json''' )
save_json(snake_case_,snake_case_,indent=snake_case_ )
print(snake_case_ )
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}_generations.txt''' ) )
if args.debug:
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}.target''' ) )
else:
shutil.rmtree(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
_A : Dict = []
for partial_result in partial_results:
records.extend(snake_case_ )
_A : Optional[Any] = sorted(snake_case_,key=lambda snake_case_ : x["id"] )
_A : List[str] = [x["""pred"""] for x in records]
return preds
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
# WAIT FOR lots of .json files
_A : Optional[Any] = time.time()
logger.info("""waiting for all nodes to finish""" )
_A : List[str] = None
while (time.time() - start_wait) < timeout:
_A : str = list(save_dir.glob("""rank_*.json""" ) )
if len(snake_case_ ) < num_replicas:
continue
try:
# make sure all json files are fully saved
_A : List[str] = lmap(snake_case_,snake_case_ )
return json_data
except JSONDecodeError:
continue
else:
raise TimeoutError("""Rank 0 gave up on waiting for other processes""" )
# Unreachable
if __name__ == "__main__":
# Usage for MT:
run_generate()
| 26 | 1 |
from __future__ import annotations
import typing
from collections.abc import Iterable
import numpy as np
_snake_case = typing.Union[Iterable[float], Iterable[int], np.ndarray] # noqa: UP007
_snake_case = typing.Union[np.floataa, int, float] # noqa: UP007
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
return np.sqrt(np.sum((np.asarray(snake_case_ ) - np.asarray(snake_case_ )) ** 2 ) )
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
return sum((va - va) ** 2 for va, va in zip(snake_case_,snake_case_ ) ) ** (1 / 2)
if __name__ == "__main__":
def lowerCAmelCase_ ( ):
from timeit import timeit
print("""Without Numpy""" )
print(
timeit(
"""euclidean_distance_no_np([1, 2, 3], [4, 5, 6])""",number=10000,globals=globals(),) )
print("""With Numpy""" )
print(
timeit(
"""euclidean_distance([1, 2, 3], [4, 5, 6])""",number=10000,globals=globals(),) )
benchmark()
| 26 |
from __future__ import annotations
import unittest
from transformers import is_tf_available
from transformers.testing_utils import require_sentencepiece, require_tf, require_tokenizers, slow
if is_tf_available():
import numpy as np
import tensorflow as tf
from transformers import TFCamembertModel
@require_tf
@require_sentencepiece
@require_tokenizers
class lowercase ( unittest.TestCase ):
@slow
def a__ ( self ) -> Any:
_A : Tuple = TFCamembertModel.from_pretrained("""jplu/tf-camembert-base""" )
_A : List[Any] = tf.convert_to_tensor(
[[5, 121, 11, 660, 16, 730, 2_5543, 110, 83, 6]] , dtype=tf.intaa , ) # J'aime le camembert !"
_A : List[str] = model(_a )["""last_hidden_state"""]
_A : Union[str, Any] = tf.TensorShape((1, 10, 768) )
self.assertEqual(output.shape , _a )
# compare the actual values for a slice.
_A : List[Any] = tf.convert_to_tensor(
[[[-0.0254, 0.0235, 0.1027], [0.0606, -0.1811, -0.0418], [-0.1561, -0.1127, 0.2687]]] , dtype=tf.floataa , )
# camembert = torch.hub.load('pytorch/fairseq', 'camembert.v0')
# camembert.eval()
# expected_slice = roberta.model.forward(input_ids)[0][:, :3, :3].detach()
self.assertTrue(np.allclose(output[:, :3, :3].numpy() , expected_slice.numpy() , atol=1e-4 ) )
| 26 | 1 |
import argparse
from collections import OrderedDict
from pathlib import Path
import torch
from transformers import (
VisualBertConfig,
VisualBertForMultipleChoice,
VisualBertForPreTraining,
VisualBertForQuestionAnswering,
VisualBertForVisualReasoning,
)
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
_snake_case = [
("bert.bert", "visual_bert"),
("bert.cls", "cls"),
("bert.classifier", "cls"),
("token_type_embeddings_visual", "visual_token_type_embeddings"),
("position_embeddings_visual", "visual_position_embeddings"),
("projection", "visual_projection"),
]
_snake_case = [
"nlvr2_coco_pre_trained.th",
"nlvr2_fine_tuned.th",
"nlvr2_pre_trained.th",
"vcr_coco_pre_train.th",
"vcr_fine_tune.th",
"vcr_pre_train.th",
"vqa_coco_pre_trained.th",
"vqa_fine_tuned.th",
"vqa_pre_trained.th",
]
def lowerCAmelCase_ ( snake_case_ ):
_A : int = torch.load(snake_case_,map_location="""cpu""" )
return sd
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_=rename_keys_prefix ):
_A : Dict = OrderedDict()
_A : Tuple = torch.arange(config.max_position_embeddings ).expand((1, -1) )
# detector_d = OrderedDict()
for key in d:
if "detector" in key:
# detector_d[key.replace('detector.','')] = d[key]
continue
_A : Dict = key
for name_pair in rename_keys_prefix:
_A : Tuple = new_key.replace(name_pair[0],name_pair[1] )
_A : Dict = d[key]
if key == "bert.cls.predictions.decoder.weight":
# Old bert code didn't have `decoder.bias`, but was added separately
_A : Optional[int] = new_d["""cls.predictions.bias"""]
return new_d
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
assert (
checkpoint_path.split("""/""" )[-1] in ACCEPTABLE_CHECKPOINTS
), f'''The checkpoint provided must be in {ACCEPTABLE_CHECKPOINTS}.'''
# Get Config
if "pre" in checkpoint_path:
_A : str = """pretraining"""
if "vcr" in checkpoint_path:
_A : int = {"""visual_embedding_dim""": 512}
elif "vqa_advanced" in checkpoint_path:
_A : Any = {"""visual_embedding_dim""": 2048}
elif "vqa" in checkpoint_path:
_A : int = {"""visual_embedding_dim""": 2048}
elif "nlvr" in checkpoint_path:
_A : Optional[int] = {"""visual_embedding_dim""": 1024}
else:
raise NotImplementedError(f'''No implementation found for `{checkpoint_path}`.''' )
else:
if "vcr" in checkpoint_path:
_A : Optional[Any] = {"""visual_embedding_dim""": 512}
_A : List[Any] = """multichoice"""
elif "vqa_advanced" in checkpoint_path:
_A : List[Any] = {"""visual_embedding_dim""": 2048}
_A : str = """vqa_advanced"""
elif "vqa" in checkpoint_path:
_A : Optional[Any] = {"""visual_embedding_dim""": 2048, """num_labels""": 3129}
_A : Any = """vqa"""
elif "nlvr" in checkpoint_path:
_A : int = {
"""visual_embedding_dim""": 1024,
"""num_labels""": 2,
}
_A : List[Any] = """nlvr"""
_A : Tuple = VisualBertConfig(**snake_case_ )
# Load State Dict
_A : Any = load_state_dict(snake_case_ )
_A : List[Any] = get_new_dict(snake_case_,snake_case_ )
if model_type == "pretraining":
_A : Tuple = VisualBertForPreTraining(snake_case_ )
elif model_type == "vqa":
_A : Tuple = VisualBertForQuestionAnswering(snake_case_ )
elif model_type == "nlvr":
_A : Union[str, Any] = VisualBertForVisualReasoning(snake_case_ )
elif model_type == "multichoice":
_A : List[Any] = VisualBertForMultipleChoice(snake_case_ )
model.load_state_dict(snake_case_ )
# Save Checkpoints
Path(snake_case_ ).mkdir(exist_ok=snake_case_ )
model.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument("orig_checkpoint_path", type=str, help="A path to .th on local filesystem.")
parser.add_argument("pytorch_dump_folder_path", type=str, help="Path to the output PyTorch model.")
_snake_case = parser.parse_args()
convert_visual_bert_checkpoint(args.orig_checkpoint_path, args.pytorch_dump_folder_path)
| 26 |
from dataclasses import dataclass
from typing import List, Optional, Union
import numpy as np
import PIL
import torch
from transformers import CLIPImageProcessor, CLIPVisionModel
from ...models import PriorTransformer
from ...pipelines import DiffusionPipeline
from ...schedulers import HeunDiscreteScheduler
from ...utils import (
BaseOutput,
is_accelerate_available,
logging,
randn_tensor,
replace_example_docstring,
)
from .renderer import ShapERenderer
_snake_case = logging.get_logger(__name__) # pylint: disable=invalid-name
_snake_case = "\n Examples:\n ```py\n >>> from PIL import Image\n >>> import torch\n >>> from diffusers import DiffusionPipeline\n >>> from diffusers.utils import export_to_gif, load_image\n\n >>> device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n\n >>> repo = \"openai/shap-e-img2img\"\n >>> pipe = DiffusionPipeline.from_pretrained(repo, torch_dtype=torch.float16)\n >>> pipe = pipe.to(device)\n\n >>> guidance_scale = 3.0\n >>> image_url = \"https://hf.co/datasets/diffusers/docs-images/resolve/main/shap-e/corgi.png\"\n >>> image = load_image(image_url).convert(\"RGB\")\n\n >>> images = pipe(\n ... image,\n ... guidance_scale=guidance_scale,\n ... num_inference_steps=64,\n ... frame_size=256,\n ... ).images\n\n >>> gif_path = export_to_gif(images[0], \"corgi_3d.gif\")\n ```\n"
@dataclass
class lowercase ( UpperCamelCase__ ):
_a = 42
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a , _a , _a , ) -> List[Any]:
super().__init__()
self.register_modules(
prior=_a , image_encoder=_a , image_processor=_a , scheduler=_a , renderer=_a , )
def a__ ( self , _a , _a , _a , _a , _a , _a ) -> str:
if latents is None:
_A : str = randn_tensor(_a , generator=_a , device=_a , dtype=_a )
else:
if latents.shape != shape:
raise ValueError(F'''Unexpected latents shape, got {latents.shape}, expected {shape}''' )
_A : Union[str, Any] = latents.to(_a )
_A : int = latents * scheduler.init_noise_sigma
return latents
def a__ ( self , _a=0 ) -> Optional[Any]:
if is_accelerate_available():
from accelerate import cpu_offload
else:
raise ImportError("""Please install accelerate via `pip install accelerate`""" )
_A : str = torch.device(F'''cuda:{gpu_id}''' )
_A : Any = [self.image_encoder, self.prior]
for cpu_offloaded_model in models:
if cpu_offloaded_model is not None:
cpu_offload(_a , _a )
@property
def a__ ( self ) -> List[Any]:
if self.device != torch.device("""meta""" ) or not hasattr(self.image_encoder , """_hf_hook""" ):
return self.device
for module in self.image_encoder.modules():
if (
hasattr(_a , """_hf_hook""" )
and hasattr(module._hf_hook , """execution_device""" )
and module._hf_hook.execution_device is not None
):
return torch.device(module._hf_hook.execution_device )
return self.device
def a__ ( self , _a , _a , _a , _a , ) -> Tuple:
if isinstance(_a , _a ) and isinstance(image[0] , torch.Tensor ):
_A : int = torch.cat(_a , axis=0 ) if image[0].ndim == 4 else torch.stack(_a , axis=0 )
if not isinstance(_a , torch.Tensor ):
_A : Dict = self.image_processor(_a , return_tensors="""pt""" ).pixel_values[0].unsqueeze(0 )
_A : int = image.to(dtype=self.image_encoder.dtype , device=_a )
_A : List[Any] = self.image_encoder(_a )["""last_hidden_state"""]
_A : List[Any] = image_embeds[:, 1:, :].contiguous() # batch_size, dim, 256
_A : Dict = image_embeds.repeat_interleave(_a , dim=0 )
if do_classifier_free_guidance:
_A : str = torch.zeros_like(_a )
# For classifier free guidance, we need to do two forward passes.
# Here we concatenate the unconditional and text embeddings into a single batch
# to avoid doing two forward passes
_A : List[str] = torch.cat([negative_image_embeds, image_embeds] )
return image_embeds
@torch.no_grad()
@replace_example_docstring(_a )
def __call__( self , _a , _a = 1 , _a = 25 , _a = None , _a = None , _a = 4.0 , _a = 64 , _a = "pil" , _a = True , ) -> Union[str, Any]:
if isinstance(_a , PIL.Image.Image ):
_A : List[Any] = 1
elif isinstance(_a , torch.Tensor ):
_A : Any = image.shape[0]
elif isinstance(_a , _a ) and isinstance(image[0] , (torch.Tensor, PIL.Image.Image) ):
_A : Union[str, Any] = len(_a )
else:
raise ValueError(
F'''`image` has to be of type `PIL.Image.Image`, `torch.Tensor`, `List[PIL.Image.Image]` or `List[torch.Tensor]` but is {type(_a )}''' )
_A : Optional[int] = self._execution_device
_A : Tuple = batch_size * num_images_per_prompt
_A : List[Any] = guidance_scale > 1.0
_A : Optional[Any] = self._encode_image(_a , _a , _a , _a )
# prior
self.scheduler.set_timesteps(_a , device=_a )
_A : Optional[int] = self.scheduler.timesteps
_A : List[str] = self.prior.config.num_embeddings
_A : int = self.prior.config.embedding_dim
_A : Optional[Any] = self.prepare_latents(
(batch_size, num_embeddings * embedding_dim) , image_embeds.dtype , _a , _a , _a , self.scheduler , )
# YiYi notes: for testing only to match ldm, we can directly create a latents with desired shape: batch_size, num_embeddings, embedding_dim
_A : List[Any] = latents.reshape(latents.shape[0] , _a , _a )
for i, t in enumerate(self.progress_bar(_a ) ):
# expand the latents if we are doing classifier free guidance
_A : List[str] = torch.cat([latents] * 2 ) if do_classifier_free_guidance else latents
_A : int = self.scheduler.scale_model_input(_a , _a )
_A : Tuple = self.prior(
_a , timestep=_a , proj_embedding=_a , ).predicted_image_embedding
# remove the variance
_A , _A : Optional[Any] = noise_pred.split(
scaled_model_input.shape[2] , dim=2 ) # batch_size, num_embeddings, embedding_dim
if do_classifier_free_guidance is not None:
_A , _A : Dict = noise_pred.chunk(2 )
_A : Tuple = noise_pred_uncond + guidance_scale * (noise_pred - noise_pred_uncond)
_A : int = self.scheduler.step(
_a , timestep=_a , sample=_a , ).prev_sample
if output_type == "latent":
return ShapEPipelineOutput(images=_a )
_A : List[str] = []
for i, latent in enumerate(_a ):
print()
_A : List[str] = self.renderer.decode(
latent[None, :] , _a , size=_a , ray_batch_size=4096 , n_coarse_samples=64 , n_fine_samples=128 , )
images.append(_a )
_A : List[Any] = torch.stack(_a )
if output_type not in ["np", "pil"]:
raise ValueError(F'''Only the output types `pil` and `np` are supported not output_type={output_type}''' )
_A : List[str] = images.cpu().numpy()
if output_type == "pil":
_A : List[Any] = [self.numpy_to_pil(_a ) for image in images]
# Offload last model to CPU
if hasattr(self , """final_offload_hook""" ) and self.final_offload_hook is not None:
self.final_offload_hook.offload()
if not return_dict:
return (images,)
return ShapEPipelineOutput(images=_a )
| 26 | 1 |
from typing import Dict, List, Optional, Tuple, Union
import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
flip_channel_order,
get_resize_output_image_size,
rescale,
resize,
to_channel_dimension_format,
)
from ...image_utils import (
ChannelDimension,
ImageInput,
PILImageResampling,
make_list_of_images,
to_numpy_array,
valid_images,
)
from ...utils import TensorType, is_torch_available, is_torch_tensor, is_vision_available, logging
if is_vision_available():
import PIL
if is_torch_available():
import torch
_snake_case = logging.get_logger(__name__)
class lowercase ( UpperCamelCase__ ):
_a = ["pixel_values"]
def __init__( self , _a = True , _a = None , _a = PILImageResampling.BILINEAR , _a = True , _a = 1 / 255 , _a = True , _a = None , _a = True , **_a , ) -> None:
super().__init__(**_a )
_A : Optional[Any] = size if size is not None else {"""shortest_edge""": 224}
_A : int = get_size_dict(_a , default_to_square=_a )
_A : Union[str, Any] = crop_size if crop_size is not None else {"""height""": 256, """width""": 256}
_A : str = get_size_dict(_a , param_name="""crop_size""" )
_A : List[Any] = do_resize
_A : Dict = size
_A : Union[str, Any] = resample
_A : int = do_rescale
_A : Any = rescale_factor
_A : Union[str, Any] = do_center_crop
_A : str = crop_size
_A : Optional[int] = do_flip_channel_order
def a__ ( self , _a , _a , _a = PIL.Image.BILINEAR , _a = None , **_a , ) -> np.ndarray:
_A : List[Any] = get_size_dict(_a , default_to_square=_a )
if "shortest_edge" not in size:
raise ValueError(F'''The `size` dictionary must contain the key `shortest_edge`. Got {size.keys()}''' )
_A : Optional[Any] = get_resize_output_image_size(_a , size=size["""shortest_edge"""] , default_to_square=_a )
return resize(_a , size=_a , resample=_a , data_format=_a , **_a )
def a__ ( self , _a , _a , _a = None , **_a , ) -> np.ndarray:
_A : Union[str, Any] = get_size_dict(_a )
if "height" not in size or "width" not in size:
raise ValueError(F'''The `size` dictionary must contain the keys `height` and `width`. Got {size.keys()}''' )
return center_crop(_a , size=(size["""height"""], size["""width"""]) , data_format=_a , **_a )
def a__ ( self , _a , _a , _a = None , **_a , ) -> List[Any]:
return rescale(_a , scale=_a , data_format=_a , **_a )
def a__ ( self , _a , _a = None ) -> np.ndarray:
return flip_channel_order(_a , data_format=_a )
def a__ ( self , _a , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = ChannelDimension.FIRST , **_a , ) -> PIL.Image.Image:
_A : int = do_resize if do_resize is not None else self.do_resize
_A : List[str] = resample if resample is not None else self.resample
_A : Union[str, Any] = do_rescale if do_rescale is not None else self.do_rescale
_A : Tuple = rescale_factor if rescale_factor is not None else self.rescale_factor
_A : Any = do_center_crop if do_center_crop is not None else self.do_center_crop
_A : List[Any] = (
do_flip_channel_order if do_flip_channel_order is not None else self.do_flip_channel_order
)
_A : Union[str, Any] = size if size is not None else self.size
_A : List[str] = get_size_dict(_a , default_to_square=_a )
_A : Any = crop_size if crop_size is not None else self.crop_size
_A : Dict = get_size_dict(_a , param_name="""crop_size""" )
_A : int = make_list_of_images(_a )
if not valid_images(_a ):
raise ValueError(
"""Invalid image type. Must be of type PIL.Image.Image, numpy.ndarray, """
"""torch.Tensor, tf.Tensor or jax.ndarray.""" )
if do_resize and size is None:
raise ValueError("""Size must be specified if do_resize is True.""" )
if do_rescale and rescale_factor is None:
raise ValueError("""Rescale factor must be specified if do_rescale is True.""" )
if do_center_crop and crop_size is None:
raise ValueError("""Crop size must be specified if do_center_crop is True.""" )
# All transformations expect numpy arrays.
_A : List[Any] = [to_numpy_array(_a ) for image in images]
if do_resize:
_A : Union[str, Any] = [self.resize(image=_a , size=_a , resample=_a ) for image in images]
if do_center_crop:
_A : str = [self.center_crop(image=_a , size=_a ) for image in images]
if do_rescale:
_A : List[Any] = [self.rescale(image=_a , scale=_a ) for image in images]
# the pretrained checkpoints assume images are BGR, not RGB
if do_flip_channel_order:
_A : Optional[Any] = [self.flip_channel_order(image=_a ) for image in images]
_A : List[str] = [to_channel_dimension_format(_a , _a ) for image in images]
_A : Union[str, Any] = {"""pixel_values""": images}
return BatchFeature(data=_a , tensor_type=_a )
def a__ ( self , _a , _a = None ) -> str:
_A : Union[str, Any] = outputs.logits
# Resize logits and compute semantic segmentation maps
if target_sizes is not None:
if len(_a ) != len(_a ):
raise ValueError(
"""Make sure that you pass in as many target sizes as the batch dimension of the logits""" )
if is_torch_tensor(_a ):
_A : Optional[int] = target_sizes.numpy()
_A : Optional[Any] = []
for idx in range(len(_a ) ):
_A : Any = torch.nn.functional.interpolate(
logits[idx].unsqueeze(dim=0 ) , size=target_sizes[idx] , mode="""bilinear""" , align_corners=_a )
_A : str = resized_logits[0].argmax(dim=0 )
semantic_segmentation.append(_a )
else:
_A : List[str] = logits.argmax(dim=1 )
_A : List[str] = [semantic_segmentation[i] for i in range(semantic_segmentation.shape[0] )]
return semantic_segmentation
| 26 |
import argparse
import collections
import json
from pathlib import Path
import requests
import torch
import yaml
from huggingface_hub import hf_hub_download
from PIL import Image
from transformers import (
MobileViTImageProcessor,
MobileViTVaConfig,
MobileViTVaForImageClassification,
MobileViTVaForSemanticSegmentation,
)
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
def lowerCAmelCase_ ( snake_case_ ):
print("""Loading config file...""" )
def flatten_yaml_as_dict(snake_case_,snake_case_="",snake_case_="." ):
_A : Union[str, Any] = []
for k, v in d.items():
_A : Optional[int] = parent_key + sep + k if parent_key else k
if isinstance(snake_case_,collections.abc.MutableMapping ):
items.extend(flatten_yaml_as_dict(snake_case_,snake_case_,sep=snake_case_ ).items() )
else:
items.append((new_key, v) )
return dict(snake_case_ )
_A : List[Any] = argparse.Namespace()
with open(snake_case_,"""r""" ) as yaml_file:
try:
_A : List[Any] = yaml.load(snake_case_,Loader=yaml.FullLoader )
_A : Optional[int] = flatten_yaml_as_dict(snake_case_ )
for k, v in flat_cfg.items():
setattr(snake_case_,snake_case_,snake_case_ )
except yaml.YAMLError as exc:
logger.error("""Error while loading config file: {}. Error message: {}""".format(snake_case_,str(snake_case_ ) ) )
return config
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Optional[Any] = MobileViTVaConfig()
_A : Tuple = False
# dataset
if task_name.startswith("""imagenet1k_""" ):
_A : Dict = 1000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : int = 384
else:
_A : int = 256
_A : List[str] = """imagenet-1k-id2label.json"""
elif task_name.startswith("""imagenet21k_to_1k_""" ):
_A : Union[str, Any] = 21000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : str = 384
else:
_A : List[Any] = 256
_A : List[str] = """imagenet-22k-id2label.json"""
elif task_name.startswith("""ade20k_""" ):
_A : int = 151
_A : int = 512
_A : Optional[int] = """ade20k-id2label.json"""
_A : Any = True
elif task_name.startswith("""voc_""" ):
_A : List[Any] = 21
_A : Dict = 512
_A : Dict = """pascal-voc-id2label.json"""
_A : int = True
# orig_config
_A : Any = load_orig_config_file(snake_case_ )
assert getattr(snake_case_,"""model.classification.name""",-1 ) == "mobilevit_v2", "Invalid model"
_A : List[Any] = getattr(snake_case_,"""model.classification.mitv2.width_multiplier""",1.0 )
assert (
getattr(snake_case_,"""model.classification.mitv2.attn_norm_layer""",-1 ) == "layer_norm_2d"
), "Norm layers other than layer_norm_2d is not supported"
_A : str = getattr(snake_case_,"""model.classification.activation.name""","""swish""" )
# config.image_size == getattr(orig_config, 'sampler.bs.crop_size_width', 256)
if is_segmentation_model:
_A : Optional[int] = getattr(snake_case_,"""model.segmentation.output_stride""",16 )
if "_deeplabv3" in task_name:
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_rates""",[12, 24, 36] )
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_out_channels""",512 )
_A : str = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_dropout""",0.1 )
# id2label
_A : List[Any] = """huggingface/label-files"""
_A : List[Any] = json.load(open(hf_hub_download(snake_case_,snake_case_,repo_type="""dataset""" ),"""r""" ) )
_A : str = {int(snake_case_ ): v for k, v in idalabel.items()}
_A : str = idalabel
_A : Dict = {v: k for k, v in idalabel.items()}
return config
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Any = dct.pop(snake_case_ )
_A : Union[str, Any] = val
def lowerCAmelCase_ ( snake_case_,snake_case_=False ):
if base_model:
_A : Optional[int] = """"""
else:
_A : Dict = """mobilevitv2."""
_A : int = []
for k in state_dict.keys():
if k[:8] == "encoder.":
_A : Any = k[8:]
else:
_A : List[str] = k
if ".block." in k:
_A : Any = k_new.replace(""".block.""",""".""" )
if ".conv." in k:
_A : List[Any] = k_new.replace(""".conv.""",""".convolution.""" )
if ".norm." in k:
_A : Any = k_new.replace(""".norm.""",""".normalization.""" )
if "conv_1." in k:
_A : int = k_new.replace("""conv_1.""",f'''{model_prefix}conv_stem.''' )
for i in [1, 2]:
if f'''layer_{i}.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.''',f'''{model_prefix}encoder.layer.{i-1}.layer.''' )
if ".exp_1x1." in k:
_A : Tuple = k_new.replace(""".exp_1x1.""",""".expand_1x1.""" )
if ".red_1x1." in k:
_A : Optional[int] = k_new.replace(""".red_1x1.""",""".reduce_1x1.""" )
for i in [3, 4, 5]:
if f'''layer_{i}.0.''' in k:
_A : Optional[int] = k_new.replace(f'''layer_{i}.0.''',f'''{model_prefix}encoder.layer.{i-1}.downsampling_layer.''' )
if f'''layer_{i}.1.local_rep.0.''' in k:
_A : Union[str, Any] = k_new.replace(f'''layer_{i}.1.local_rep.0.''',f'''{model_prefix}encoder.layer.{i-1}.conv_kxk.''' )
if f'''layer_{i}.1.local_rep.1.''' in k:
_A : str = k_new.replace(f'''layer_{i}.1.local_rep.1.''',f'''{model_prefix}encoder.layer.{i-1}.conv_1x1.''' )
for i in [3, 4, 5]:
if i == 3:
_A : Optional[int] = [0, 1]
elif i == 4:
_A : Union[str, Any] = [0, 1, 2, 3]
elif i == 5:
_A : Optional[Any] = [0, 1, 2]
for j in j_in:
if f'''layer_{i}.1.global_rep.{j}.''' in k:
_A : Union[str, Any] = k_new.replace(
f'''layer_{i}.1.global_rep.{j}.''',f'''{model_prefix}encoder.layer.{i-1}.transformer.layer.{j}.''' )
if f'''layer_{i}.1.global_rep.{j+1}.''' in k:
_A : List[str] = k_new.replace(
f'''layer_{i}.1.global_rep.{j+1}.''',f'''{model_prefix}encoder.layer.{i-1}.layernorm.''' )
if f'''layer_{i}.1.conv_proj.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.1.conv_proj.''',f'''{model_prefix}encoder.layer.{i-1}.conv_projection.''' )
if "pre_norm_attn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_attn.0.""","""layernorm_before.""" )
if "pre_norm_attn.1." in k:
_A : str = k_new.replace("""pre_norm_attn.1.""","""attention.""" )
if "pre_norm_ffn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_ffn.0.""","""layernorm_after.""" )
if "pre_norm_ffn.1." in k:
_A : Dict = k_new.replace("""pre_norm_ffn.1.""","""ffn.conv1.""" )
if "pre_norm_ffn.3." in k:
_A : List[str] = k_new.replace("""pre_norm_ffn.3.""","""ffn.conv2.""" )
if "classifier.1." in k:
_A : List[str] = k_new.replace("""classifier.1.""","""classifier.""" )
if "seg_head." in k:
_A : List[Any] = k_new.replace("""seg_head.""","""segmentation_head.""" )
if ".aspp_layer." in k:
_A : List[Any] = k_new.replace(""".aspp_layer.""",""".""" )
if ".aspp_pool." in k:
_A : Optional[Any] = k_new.replace(""".aspp_pool.""",""".""" )
rename_keys.append((k, k_new) )
return rename_keys
def lowerCAmelCase_ ( snake_case_ ):
_A : Tuple = []
for k in state_dict.keys():
if k.startswith("""seg_head.aux_head.""" ):
keys_to_ignore.append(snake_case_ )
for k in keys_to_ignore:
state_dict.pop(snake_case_,snake_case_ )
def lowerCAmelCase_ ( ):
_A : Dict = """http://images.cocodataset.org/val2017/000000039769.jpg"""
# url = "https://cdn.britannica.com/86/141086-050-9D7C75EE/Gulfstream-G450-business-jet-passengers.jpg"
_A : List[Any] = Image.open(requests.get(snake_case_,stream=snake_case_ ).raw )
return im
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : List[Any] = get_mobilevitva_config(snake_case_,snake_case_ )
# load original state_dict
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
# load huggingface model
if task_name.startswith("""ade20k_""" ) or task_name.startswith("""voc_""" ):
_A : Optional[Any] = MobileViTVaForSemanticSegmentation(snake_case_ ).eval()
_A : str = False
else:
_A : int = MobileViTVaForImageClassification(snake_case_ ).eval()
_A : List[Any] = False
# remove and rename some keys of load the original model
_A : List[Any] = checkpoint
remove_unused_keys(snake_case_ )
_A : Optional[Any] = create_rename_keys(snake_case_,base_model=snake_case_ )
for rename_key_src, rename_key_dest in rename_keys:
rename_key(snake_case_,snake_case_,snake_case_ )
# load modified state_dict
model.load_state_dict(snake_case_ )
# Check outputs on an image, prepared by MobileViTImageProcessor
_A : str = MobileViTImageProcessor(crop_size=config.image_size,size=config.image_size + 32 )
_A : List[Any] = image_processor(images=prepare_img(),return_tensors="""pt""" )
_A : Optional[Any] = model(**snake_case_ )
# verify classification model
if task_name.startswith("""imagenet""" ):
_A : List[Any] = outputs.logits
_A : Optional[int] = logits.argmax(-1 ).item()
print("""Predicted class:""",model.config.idalabel[predicted_class_idx] )
if task_name.startswith("""imagenet1k_256""" ) and config.width_multiplier == 1.0:
# expected_logits for base variant
_A : int = torch.tensor([-1.63_36e00, -7.32_04e-02, -5.18_83e-01] )
assert torch.allclose(logits[0, :3],snake_case_,atol=1e-4 )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ )
print(f'''Saving model {task_name} to {pytorch_dump_folder_path}''' )
model.save_pretrained(snake_case_ )
print(f'''Saving image processor to {pytorch_dump_folder_path}''' )
image_processor.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
"--task",
default="imagenet1k_256",
type=str,
help=(
"Name of the task for which the MobileViTV2 model you'd like to convert is trained on . "
"\n Classification (ImageNet-1k)\n - MobileViTV2 (256x256) : imagenet1k_256\n - MobileViTV2 (Trained on 256x256 and Finetuned on 384x384) : imagenet1k_384\n - MobileViTV2 (Trained on ImageNet-21k and Finetuned on ImageNet-1k 256x256) :\n imagenet21k_to_1k_256\n - MobileViTV2 (Trained on ImageNet-21k, Finetuned on ImageNet-1k 256x256, and Finetuned on\n ImageNet-1k 384x384) : imagenet21k_to_1k_384\n Segmentation\n - ADE20K Dataset : ade20k_deeplabv3\n - Pascal VOC 2012 Dataset: voc_deeplabv3\n "
),
choices=[
"imagenet1k_256",
"imagenet1k_384",
"imagenet21k_to_1k_256",
"imagenet21k_to_1k_384",
"ade20k_deeplabv3",
"voc_deeplabv3",
],
)
parser.add_argument(
"--orig_checkpoint_path", required=True, type=str, help="Path to the original state dict (.pt file)."
)
parser.add_argument("--orig_config_path", required=True, type=str, help="Path to the original config file.")
parser.add_argument(
"--pytorch_dump_folder_path", required=True, type=str, help="Path to the output PyTorch model directory."
)
_snake_case = parser.parse_args()
convert_mobilevitva_checkpoint(
args.task, args.orig_checkpoint_path, args.orig_config_path, args.pytorch_dump_folder_path
)
| 26 | 1 |
from collections.abc import Iterable
from typing import Any
class lowercase :
def __init__( self , _a = None ) -> Dict:
_A : Any = value
_A : Node | None = None # Added in order to delete a node easier
_A : Node | None = None
_A : Node | None = None
def __repr__( self ) -> str:
from pprint import pformat
if self.left is None and self.right is None:
return str(self.value )
return pformat({F'''{self.value}''': (self.left, self.right)} , indent=1 )
class lowercase :
def __init__( self , _a = None ) -> Optional[Any]:
_A : Dict = root
def __str__( self ) -> str:
return str(self.root )
def a__ ( self , _a , _a ) -> None:
if new_children is not None: # reset its kids
_A : Tuple = node.parent
if node.parent is not None: # reset its parent
if self.is_right(_a ): # If it is the right children
_A : Dict = new_children
else:
_A : int = new_children
else:
_A : List[str] = new_children
def a__ ( self , _a ) -> bool:
if node.parent and node.parent.right:
return node == node.parent.right
return False
def a__ ( self ) -> bool:
return self.root is None
def a__ ( self , _a ) -> None:
_A : List[Any] = Node(_a ) # create a new Node
if self.empty(): # if Tree is empty
_A : Union[str, Any] = new_node # set its root
else: # Tree is not empty
_A : int = self.root # from root
if parent_node is None:
return
while True: # While we don't get to a leaf
if value < parent_node.value: # We go left
if parent_node.left is None:
_A : List[Any] = new_node # We insert the new node in a leaf
break
else:
_A : Union[str, Any] = parent_node.left
else:
if parent_node.right is None:
_A : List[Any] = new_node
break
else:
_A : int = parent_node.right
_A : Optional[int] = parent_node
def a__ ( self , *_a ) -> None:
for value in values:
self.__insert(_a )
def a__ ( self , _a ) -> Node | None:
if self.empty():
raise IndexError("""Warning: Tree is empty! please use another.""" )
else:
_A : Union[str, Any] = self.root
# use lazy evaluation here to avoid NoneType Attribute error
while node is not None and node.value is not value:
_A : int = node.left if value < node.value else node.right
return node
def a__ ( self , _a = None ) -> Node | None:
if node is None:
if self.root is None:
return None
_A : Union[str, Any] = self.root
if not self.empty():
while node.right is not None:
_A : str = node.right
return node
def a__ ( self , _a = None ) -> Node | None:
if node is None:
_A : Dict = self.root
if self.root is None:
return None
if not self.empty():
_A : str = self.root
while node.left is not None:
_A : int = node.left
return node
def a__ ( self , _a ) -> None:
_A : Tuple = self.search(_a ) # Look for the node with that label
if node is not None:
if node.left is None and node.right is None: # If it has no children
self.__reassign_nodes(_a , _a )
elif node.left is None: # Has only right children
self.__reassign_nodes(_a , node.right )
elif node.right is None: # Has only left children
self.__reassign_nodes(_a , node.left )
else:
_A : Tuple = self.get_max(
node.left ) # Gets the max value of the left branch
self.remove(tmp_node.value ) # type: ignore
_A : Union[str, Any] = (
tmp_node.value # type: ignore
) # Assigns the value to the node to delete and keep tree structure
def a__ ( self , _a ) -> Iterable:
if node is not None:
yield node # Preorder Traversal
yield from self.preorder_traverse(node.left )
yield from self.preorder_traverse(node.right )
def a__ ( self , _a=None ) -> Any:
if traversal_function is None:
return self.preorder_traverse(self.root )
else:
return traversal_function(self.root )
def a__ ( self , _a , _a ) -> None:
if node:
self.inorder(_a , node.left )
arr.append(node.value )
self.inorder(_a , node.right )
def a__ ( self , _a , _a ) -> int:
_A : list[int] = []
self.inorder(_a , _a ) # append all values to list using inorder traversal
return arr[k - 1]
def lowerCAmelCase_ ( snake_case_ ):
_A : List[Any] = []
if curr_node is not None:
_A : Union[str, Any] = postorder(curr_node.left ) + postorder(curr_node.right ) + [curr_node]
return node_list
def lowerCAmelCase_ ( ):
_A : int = (8, 3, 6, 1, 10, 14, 13, 4, 7)
_A : List[Any] = BinarySearchTree()
for i in testlist:
t.insert(snake_case_ )
# Prints all the elements of the list in order traversal
print(snake_case_ )
if t.search(6 ) is not None:
print("""The value 6 exists""" )
else:
print("""The value 6 doesn't exist""" )
if t.search(-1 ) is not None:
print("""The value -1 exists""" )
else:
print("""The value -1 doesn't exist""" )
if not t.empty():
print("""Max Value: """,t.get_max().value ) # type: ignore
print("""Min Value: """,t.get_min().value ) # type: ignore
for i in testlist:
t.remove(snake_case_ )
print(snake_case_ )
if __name__ == "__main__":
import doctest
doctest.testmod(verbose=True)
| 26 |
import torch
from diffusers import DPMSolverSDEScheduler
from diffusers.utils import torch_device
from diffusers.utils.testing_utils import require_torchsde
from .test_schedulers import SchedulerCommonTest
@require_torchsde
class lowercase ( UpperCamelCase__ ):
_a = (DPMSolverSDEScheduler,)
_a = 1_0
def a__ ( self , **_a ) -> Optional[Any]:
_A : str = {
"""num_train_timesteps""": 1100,
"""beta_start""": 0.0001,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
"""noise_sampler_seed""": 0,
}
config.update(**_a )
return config
def a__ ( self ) -> Tuple:
for timesteps in [10, 50, 100, 1000]:
self.check_over_configs(num_train_timesteps=_a )
def a__ ( self ) -> Optional[int]:
for beta_start, beta_end in zip([0.00001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ):
self.check_over_configs(beta_start=_a , beta_end=_a )
def a__ ( self ) -> Any:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=_a )
def a__ ( self ) -> Optional[int]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=_a )
def a__ ( self ) -> Optional[int]:
_A : Any = self.scheduler_classes[0]
_A : List[str] = self.get_scheduler_config()
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Dict = self.dummy_model()
_A : Any = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Dict = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : str = model(_a , _a )
_A : List[Any] = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Dict = torch.sum(torch.abs(_a ) )
_A : Dict = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.47821044921875 ) < 1e-2
assert abs(result_mean.item() - 0.2178705964565277 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59352111816406 ) < 1e-2
assert abs(result_mean.item() - 0.22342906892299652 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Optional[Any]:
_A : Dict = self.scheduler_classes[0]
_A : Optional[int] = self.get_scheduler_config(prediction_type="""v_prediction""" )
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Tuple = self.dummy_model()
_A : int = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Tuple = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : int = scheduler.scale_model_input(_a , _a )
_A : Tuple = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Optional[Any] = torch.sum(torch.abs(_a ) )
_A : List[Any] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 124.77149200439453 ) < 1e-2
assert abs(result_mean.item() - 0.16226289014816284 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 128.1663360595703 ) < 1e-2
assert abs(result_mean.item() - 0.16688326001167297 ) < 1e-3
else:
assert abs(result_sum.item() - 119.8487548828125 ) < 1e-2
assert abs(result_mean.item() - 0.1560530662536621 ) < 1e-3
def a__ ( self ) -> List[str]:
_A : Union[str, Any] = self.scheduler_classes[0]
_A : List[Any] = self.get_scheduler_config()
_A : List[str] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Union[str, Any] = self.dummy_model()
_A : Optional[Any] = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
_A : int = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Dict = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : str = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.46957397460938 ) < 1e-2
assert abs(result_mean.item() - 0.21805934607982635 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59353637695312 ) < 1e-2
assert abs(result_mean.item() - 0.22342908382415771 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Union[str, Any]:
_A : List[Any] = self.scheduler_classes[0]
_A : Optional[Any] = self.get_scheduler_config()
_A : int = scheduler_class(**_a , use_karras_sigmas=_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Optional[Any] = self.dummy_model()
_A : Dict = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
_A : str = sample.to(_a )
for t in scheduler.timesteps:
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : List[str] = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : List[str] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 176.66974135742188 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 177.63653564453125 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
else:
assert abs(result_sum.item() - 170.3135223388672 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
| 26 | 1 |
# Copyright 2023 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from ..models.whisper import WhisperForConditionalGeneration, WhisperProcessor
from .base import PipelineTool
class lowercase ( UpperCamelCase__ ):
_a = "openai/whisper-base"
_a = (
"This is a tool that transcribes an audio into text. It takes an input named `audio` and returns the "
"transcribed text."
)
_a = "transcriber"
_a = WhisperProcessor
_a = WhisperForConditionalGeneration
_a = ["audio"]
_a = ["text"]
def a__ ( self , _a ) -> Dict:
return self.pre_processor(_a , return_tensors="""pt""" ).input_features
def a__ ( self , _a ) -> str:
return self.model.generate(inputs=_a )
def a__ ( self , _a ) -> Tuple:
return self.pre_processor.batch_decode(_a , skip_special_tokens=_a )[0]
| 26 |
# DISCLAIMER: This file is strongly influenced by https://github.com/yang-song/score_sde_pytorch
import math
from typing import Union
import torch
from ..configuration_utils import ConfigMixin, register_to_config
from ..utils import randn_tensor
from .scheduling_utils import SchedulerMixin
class lowercase ( UpperCamelCase__,UpperCamelCase__ ):
_a = 1
@register_to_config
def __init__( self , _a=2000 , _a=0.1 , _a=20 , _a=1e-3 ) -> List[Any]:
_A : Dict = None
_A : List[Any] = None
_A : Dict = None
def a__ ( self , _a , _a = None ) -> Union[str, Any]:
_A : Union[str, Any] = torch.linspace(1 , self.config.sampling_eps , _a , device=_a )
def a__ ( self , _a , _a , _a , _a=None ) -> Dict:
if self.timesteps is None:
raise ValueError(
"""`self.timesteps` is not set, you need to run 'set_timesteps' after creating the scheduler""" )
# TODO(Patrick) better comments + non-PyTorch
# postprocess model score
_A : Any = (
-0.25 * t**2 * (self.config.beta_max - self.config.beta_min) - 0.5 * t * self.config.beta_min
)
_A : List[Any] = torch.sqrt(1.0 - torch.exp(2.0 * log_mean_coeff ) )
_A : List[str] = std.flatten()
while len(std.shape ) < len(score.shape ):
_A : List[Any] = std.unsqueeze(-1 )
_A : int = -score / std
# compute
_A : Tuple = -1.0 / len(self.timesteps )
_A : str = self.config.beta_min + t * (self.config.beta_max - self.config.beta_min)
_A : List[str] = beta_t.flatten()
while len(beta_t.shape ) < len(x.shape ):
_A : Union[str, Any] = beta_t.unsqueeze(-1 )
_A : Tuple = -0.5 * beta_t * x
_A : Tuple = torch.sqrt(_a )
_A : Dict = drift - diffusion**2 * score
_A : Dict = x + drift * dt
# add noise
_A : Any = randn_tensor(x.shape , layout=x.layout , generator=_a , device=x.device , dtype=x.dtype )
_A : str = x_mean + diffusion * math.sqrt(-dt ) * noise
return x, x_mean
def __len__( self ) -> Optional[Any]:
return self.config.num_train_timesteps
| 26 | 1 |
from unittest.mock import Mock, patch
from file_transfer.send_file import send_file
@patch("""socket.socket""" )
@patch("""builtins.open""" )
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
# ===== initialization =====
_A : List[str] = Mock()
_A : Optional[int] = conn, Mock()
_A : Union[str, Any] = iter([1, None] )
_A : List[str] = lambda snake_case_ : next(snake_case_ )
# ===== invoke =====
send_file(filename="""mytext.txt""",testing=snake_case_ )
# ===== ensurance =====
sock.assert_called_once()
sock.return_value.bind.assert_called_once()
sock.return_value.listen.assert_called_once()
sock.return_value.accept.assert_called_once()
conn.recv.assert_called_once()
file.return_value.__enter__.assert_called_once()
file.return_value.__enter__.return_value.read.assert_called()
conn.send.assert_called_once()
conn.close.assert_called_once()
sock.return_value.shutdown.assert_called_once()
sock.return_value.close.assert_called_once()
| 26 |
import os
from shutil import copyfile
from typing import List, Optional, Tuple
from ...tokenization_utils import AddedToken
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import is_sentencepiece_available, logging
if is_sentencepiece_available():
from .tokenization_fnet import FNetTokenizer
else:
_snake_case = None
_snake_case = logging.get_logger(__name__)
_snake_case = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.json"}
_snake_case = {
"vocab_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/spiece.model",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/spiece.model",
},
"tokenizer_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/tokenizer.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/tokenizer.json",
},
}
_snake_case = {
"google/fnet-base": 512,
"google/fnet-large": 512,
}
_snake_case = "▁"
class lowercase ( UpperCamelCase__ ):
_a = VOCAB_FILES_NAMES
_a = PRETRAINED_VOCAB_FILES_MAP
_a = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
_a = ["input_ids", "token_type_ids"]
_a = FNetTokenizer
def __init__( self , _a=None , _a=None , _a=False , _a=True , _a=True , _a="<unk>" , _a="[SEP]" , _a="<pad>" , _a="[CLS]" , _a="[MASK]" , **_a , ) -> Optional[int]:
# Mask token behave like a normal word, i.e. include the space before it and
# is included in the raw text, there should be a match in a non-normalized sentence.
_A : int = (
AddedToken(_a , lstrip=_a , rstrip=_a , normalized=_a )
if isinstance(_a , _a )
else mask_token
)
super().__init__(
_a , tokenizer_file=_a , do_lower_case=_a , remove_space=_a , keep_accents=_a , unk_token=_a , sep_token=_a , pad_token=_a , cls_token=_a , mask_token=_a , **_a , )
_A : Optional[int] = do_lower_case
_A : List[Any] = remove_space
_A : str = keep_accents
_A : int = vocab_file
_A : int = False if not self.vocab_file else True
def a__ ( self , _a , _a = None ) -> List[int]:
_A : str = [self.sep_token_id]
_A : Dict = [self.cls_token_id]
if token_ids_a is None:
return cls + token_ids_a + sep
return cls + token_ids_a + sep + token_ids_a + sep
def a__ ( self , _a , _a = None ) -> List[int]:
_A : Any = [self.sep_token_id]
_A : Any = [self.cls_token_id]
if token_ids_a is None:
return len(cls + token_ids_a + sep ) * [0]
return len(cls + token_ids_a + sep ) * [0] + len(token_ids_a + sep ) * [1]
def a__ ( self , _a , _a = None ) -> Tuple[str]:
if not os.path.isdir(_a ):
logger.error(F'''Vocabulary path ({save_directory}) should be a directory''' )
return
_A : List[str] = os.path.join(
_a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
if os.path.abspath(self.vocab_file ) != os.path.abspath(_a ):
copyfile(self.vocab_file , _a )
return (out_vocab_file,)
| 26 | 1 |
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_flax_available,
is_tf_available,
is_tokenizers_available,
is_torch_available,
)
_snake_case = {
"configuration_roberta": ["ROBERTA_PRETRAINED_CONFIG_ARCHIVE_MAP", "RobertaConfig", "RobertaOnnxConfig"],
"tokenization_roberta": ["RobertaTokenizer"],
}
try:
if not is_tokenizers_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = ["RobertaTokenizerFast"]
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST",
"RobertaForCausalLM",
"RobertaForMaskedLM",
"RobertaForMultipleChoice",
"RobertaForQuestionAnswering",
"RobertaForSequenceClassification",
"RobertaForTokenClassification",
"RobertaModel",
"RobertaPreTrainedModel",
]
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"TF_ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST",
"TFRobertaForCausalLM",
"TFRobertaForMaskedLM",
"TFRobertaForMultipleChoice",
"TFRobertaForQuestionAnswering",
"TFRobertaForSequenceClassification",
"TFRobertaForTokenClassification",
"TFRobertaMainLayer",
"TFRobertaModel",
"TFRobertaPreTrainedModel",
]
try:
if not is_flax_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"FlaxRobertaForCausalLM",
"FlaxRobertaForMaskedLM",
"FlaxRobertaForMultipleChoice",
"FlaxRobertaForQuestionAnswering",
"FlaxRobertaForSequenceClassification",
"FlaxRobertaForTokenClassification",
"FlaxRobertaModel",
"FlaxRobertaPreTrainedModel",
]
if TYPE_CHECKING:
from .configuration_roberta import ROBERTA_PRETRAINED_CONFIG_ARCHIVE_MAP, RobertaConfig, RobertaOnnxConfig
from .tokenization_roberta import RobertaTokenizer
try:
if not is_tokenizers_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .tokenization_roberta_fast import RobertaTokenizerFast
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_roberta import (
ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST,
RobertaForCausalLM,
RobertaForMaskedLM,
RobertaForMultipleChoice,
RobertaForQuestionAnswering,
RobertaForSequenceClassification,
RobertaForTokenClassification,
RobertaModel,
RobertaPreTrainedModel,
)
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_tf_roberta import (
TF_ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST,
TFRobertaForCausalLM,
TFRobertaForMaskedLM,
TFRobertaForMultipleChoice,
TFRobertaForQuestionAnswering,
TFRobertaForSequenceClassification,
TFRobertaForTokenClassification,
TFRobertaMainLayer,
TFRobertaModel,
TFRobertaPreTrainedModel,
)
try:
if not is_flax_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_flax_roberta import (
FlaxRobertaForCausalLM,
FlaxRobertaForMaskedLM,
FlaxRobertaForMultipleChoice,
FlaxRobertaForQuestionAnswering,
FlaxRobertaForSequenceClassification,
FlaxRobertaForTokenClassification,
FlaxRobertaModel,
FlaxRobertaPreTrainedModel,
)
else:
import sys
_snake_case = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
| 26 |
from math import asin, atan, cos, radians, sin, sqrt, tan
_snake_case = 6_3_7_8_1_3_7.0
_snake_case = 6_3_5_6_7_5_2.3_1_4_2_4_5
_snake_case = 6378137
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : Any = (AXIS_A - AXIS_B) / AXIS_A
_A : Optional[int] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : List[str] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : Optional[Any] = radians(snake_case_ )
_A : str = radians(snake_case_ )
# Equation
_A : Dict = sin((phi_a - phi_a) / 2 )
_A : List[str] = sin((lambda_a - lambda_a) / 2 )
# Square both values
sin_sq_phi *= sin_sq_phi
sin_sq_lambda *= sin_sq_lambda
_A : Optional[int] = sqrt(sin_sq_phi + (cos(snake_case_ ) * cos(snake_case_ ) * sin_sq_lambda) )
return 2 * RADIUS * asin(snake_case_ )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
from __future__ import annotations
import unittest
from transformers import is_tf_available
from transformers.testing_utils import require_sentencepiece, require_tf, require_tokenizers, slow
if is_tf_available():
import numpy as np
import tensorflow as tf
from transformers import TFCamembertModel
@require_tf
@require_sentencepiece
@require_tokenizers
class lowercase ( unittest.TestCase ):
@slow
def a__ ( self ) -> Any:
_A : Tuple = TFCamembertModel.from_pretrained("""jplu/tf-camembert-base""" )
_A : List[Any] = tf.convert_to_tensor(
[[5, 121, 11, 660, 16, 730, 2_5543, 110, 83, 6]] , dtype=tf.intaa , ) # J'aime le camembert !"
_A : List[str] = model(_a )["""last_hidden_state"""]
_A : Union[str, Any] = tf.TensorShape((1, 10, 768) )
self.assertEqual(output.shape , _a )
# compare the actual values for a slice.
_A : List[Any] = tf.convert_to_tensor(
[[[-0.0254, 0.0235, 0.1027], [0.0606, -0.1811, -0.0418], [-0.1561, -0.1127, 0.2687]]] , dtype=tf.floataa , )
# camembert = torch.hub.load('pytorch/fairseq', 'camembert.v0')
# camembert.eval()
# expected_slice = roberta.model.forward(input_ids)[0][:, :3, :3].detach()
self.assertTrue(np.allclose(output[:, :3, :3].numpy() , expected_slice.numpy() , atol=1e-4 ) )
| 26 |
from __future__ import absolute_import, division, print_function, unicode_literals
from torch import nn
from torch.nn import CrossEntropyLoss, MSELoss
from transformers import RobertaConfig
from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
from transformers.models.roberta.modeling_roberta import (
ROBERTA_INPUTS_DOCSTRING,
ROBERTA_START_DOCSTRING,
RobertaEmbeddings,
)
from .modeling_highway_bert import BertPreTrainedModel, DeeBertModel, HighwayException, entropy
@add_start_docstrings(
"The RoBERTa Model transformer with early exiting (DeeRoBERTa). ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> Optional[int]:
super().__init__(_a )
_A : Union[str, Any] = RobertaEmbeddings(_a )
self.init_weights()
@add_start_docstrings(
"RoBERTa Model (with early exiting - DeeRoBERTa) with a classifier on top,\n also takes care of multi-layer training. ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> str:
super().__init__(_a )
_A : Any = config.num_labels
_A : Dict = config.num_hidden_layers
_A : List[str] = DeeRobertaModel(_a )
_A : int = nn.Dropout(config.hidden_dropout_prob )
_A : int = nn.Linear(config.hidden_size , self.config.num_labels )
@add_start_docstrings_to_model_forward(_a )
def a__ ( self , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=-1 , _a=False , ) -> Any:
_A : Optional[int] = self.num_layers
try:
_A : List[str] = self.roberta(
_a , attention_mask=_a , token_type_ids=_a , position_ids=_a , head_mask=_a , inputs_embeds=_a , )
_A : List[str] = outputs[1]
_A : List[str] = self.dropout(_a )
_A : Optional[Any] = self.classifier(_a )
_A : List[Any] = (logits,) + outputs[2:] # add hidden states and attention if they are here
except HighwayException as e:
_A : List[Any] = e.message
_A : Optional[int] = e.exit_layer
_A : Optional[int] = outputs[0]
if not self.training:
_A : int = entropy(_a )
_A : int = []
_A : int = []
if labels is not None:
if self.num_labels == 1:
# We are doing regression
_A : Union[str, Any] = MSELoss()
_A : Tuple = loss_fct(logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Dict = loss_fct(logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
# work with highway exits
_A : Optional[Any] = []
for highway_exit in outputs[-1]:
_A : Tuple = highway_exit[0]
if not self.training:
highway_logits_all.append(_a )
highway_entropy.append(highway_exit[2] )
if self.num_labels == 1:
# We are doing regression
_A : List[str] = MSELoss()
_A : Optional[int] = loss_fct(highway_logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Tuple = loss_fct(highway_logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
highway_losses.append(_a )
if train_highway:
_A : Dict = (sum(highway_losses[:-1] ),) + outputs
# exclude the final highway, of course
else:
_A : int = (loss,) + outputs
if not self.training:
_A : Optional[Any] = outputs + ((original_entropy, highway_entropy), exit_layer)
if output_layer >= 0:
_A : Union[str, Any] = (
(outputs[0],) + (highway_logits_all[output_layer],) + outputs[2:]
) # use the highway of the last layer
return outputs # (loss), logits, (hidden_states), (attentions), entropy
| 26 | 1 |
from manim import *
class lowercase ( UpperCamelCase__ ):
def a__ ( self ) -> Dict:
_A : List[Any] = Rectangle(height=0.5 , width=0.5 )
_A : Optional[int] = Rectangle(height=0.46 , width=0.46 ).set_stroke(width=0 )
_A : Optional[Any] = [mem.copy() for i in range(6 )]
_A : List[Any] = [mem.copy() for i in range(6 )]
_A : Optional[Any] = VGroup(*_a ).arrange(_a , buff=0 )
_A : Any = VGroup(*_a ).arrange(_a , buff=0 )
_A : Optional[int] = VGroup(_a , _a ).arrange(_a , buff=0 )
_A : Optional[Any] = Text("""CPU""" , font_size=24 )
_A : List[str] = Group(_a , _a ).arrange(_a , buff=0.5 , aligned_edge=_a )
cpu.move_to([-2.5, -0.5, 0] )
self.add(_a )
_A : Optional[Any] = [mem.copy() for i in range(4 )]
_A : Optional[int] = VGroup(*_a ).arrange(_a , buff=0 )
_A : Optional[Any] = Text("""GPU""" , font_size=24 )
_A : Union[str, Any] = Group(_a , _a ).arrange(_a , buff=0.5 , aligned_edge=_a )
gpu.move_to([-1, -1, 0] )
self.add(_a )
_A : str = [mem.copy() for i in range(6 )]
_A : int = VGroup(*_a ).arrange(_a , buff=0 )
_A : str = Text("""Model""" , font_size=24 )
_A : List[Any] = Group(_a , _a ).arrange(_a , buff=0.5 , aligned_edge=_a )
model.move_to([3, -1.0, 0] )
self.add(_a )
_A : List[str] = []
for i, rect in enumerate(_a ):
rect.set_stroke(_a )
# target = fill.copy().set_fill(YELLOW, opacity=0.7)
# target.move_to(rect)
# self.add(target)
_A : str = Rectangle(height=0.46 / 4 , width=0.46 / 3 ).set_stroke(width=0.0 ).set_fill(_a , opacity=0.7 )
if i == 0:
cpu_target.next_to(cpu_left_col_base[0].get_corner(DOWN + LEFT ) , buff=0.02 , direction=_a )
cpu_target.set_x(cpu_target.get_x() + 0.1 )
elif i == 3:
cpu_target.next_to(cpu_targs[0] , direction=_a , buff=0.0 )
else:
cpu_target.next_to(cpu_targs[i - 1] , direction=_a , buff=0.0 )
self.add(_a )
cpu_targs.append(_a )
_A : Union[str, Any] = [mem.copy() for i in range(6 )]
_A : str = VGroup(*_a ).arrange(_a , buff=0 )
_A : List[str] = Text("""Loaded Checkpoint""" , font_size=24 )
_A : Optional[int] = Group(_a , _a ).arrange(_a , aligned_edge=_a , buff=0.4 )
checkpoint.move_to([3, 0.5, 0] )
_A : List[Any] = Square(side_length=2.2 )
key.move_to([-5, 2, 0] )
_A : int = MarkupText(
F'''<b>Key:</b>\n\n<span fgcolor=\'{YELLOW}\'>●</span> Empty Model''' , font_size=18 , )
key_text.move_to([-5, 2.4, 0] )
self.add(_a , _a )
_A : List[str] = MarkupText(
F'''<span fgcolor=\'{BLUE}\'>●</span> Checkpoint''' , font_size=18 , )
blue_text.next_to(_a , DOWN * 2.4 , aligned_edge=key_text.get_left() )
_A : Optional[int] = MarkupText(
F'''Next, a <i><span fgcolor="{BLUE}">second</span></i> model is loaded into memory,\nwith the weights of a <span fgcolor="{BLUE}">single shard</span>.''' , font_size=24 , )
step_a.move_to([2, 2, 0] )
self.play(Write(_a ) , Write(_a ) )
self.play(Write(_a , run_time=1 ) , Create(_a , run_time=1 ) )
_A : Optional[int] = []
_A : Dict = []
for i, rect in enumerate(_a ):
_A : int = fill.copy().set_fill(_a , opacity=0.7 )
target.move_to(_a )
first_animations.append(GrowFromCenter(_a , run_time=1 ) )
_A : Optional[Any] = target.copy()
cpu_target.generate_target()
if i < 5:
cpu_target.target.move_to(cpu_left_col_base[i + 1] )
else:
cpu_target.target.move_to(cpu_right_col_base[i - 5] )
second_animations.append(MoveToTarget(_a , run_time=1.5 ) )
self.play(*_a )
self.play(*_a )
self.wait()
| 26 |
from collections import OrderedDict
from typing import Mapping
from ...configuration_utils import PretrainedConfig
from ...onnx import OnnxConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"facebook/xmod-base": "https://huggingface.co/facebook/xmod-base/resolve/main/config.json",
"facebook/xmod-large-prenorm": "https://huggingface.co/facebook/xmod-large-prenorm/resolve/main/config.json",
"facebook/xmod-base-13-125k": "https://huggingface.co/facebook/xmod-base-13-125k/resolve/main/config.json",
"facebook/xmod-base-30-125k": "https://huggingface.co/facebook/xmod-base-30-125k/resolve/main/config.json",
"facebook/xmod-base-30-195k": "https://huggingface.co/facebook/xmod-base-30-195k/resolve/main/config.json",
"facebook/xmod-base-60-125k": "https://huggingface.co/facebook/xmod-base-60-125k/resolve/main/config.json",
"facebook/xmod-base-60-265k": "https://huggingface.co/facebook/xmod-base-60-265k/resolve/main/config.json",
"facebook/xmod-base-75-125k": "https://huggingface.co/facebook/xmod-base-75-125k/resolve/main/config.json",
"facebook/xmod-base-75-269k": "https://huggingface.co/facebook/xmod-base-75-269k/resolve/main/config.json",
}
class lowercase ( UpperCamelCase__ ):
_a = "xmod"
def __init__( self , _a=3_0522 , _a=768 , _a=12 , _a=12 , _a=3072 , _a="gelu" , _a=0.1 , _a=0.1 , _a=512 , _a=2 , _a=0.02 , _a=1e-12 , _a=1 , _a=0 , _a=2 , _a="absolute" , _a=True , _a=None , _a=False , _a=2 , _a=False , _a=True , _a=True , _a=("en_XX",) , _a=None , **_a , ) -> str:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Tuple = vocab_size
_A : Union[str, Any] = hidden_size
_A : Dict = num_hidden_layers
_A : Dict = num_attention_heads
_A : List[Any] = hidden_act
_A : Optional[Any] = intermediate_size
_A : Any = hidden_dropout_prob
_A : str = attention_probs_dropout_prob
_A : Dict = max_position_embeddings
_A : Any = type_vocab_size
_A : List[Any] = initializer_range
_A : int = layer_norm_eps
_A : int = position_embedding_type
_A : Any = use_cache
_A : int = classifier_dropout
_A : int = pre_norm
_A : Optional[Any] = adapter_reduction_factor
_A : List[Any] = adapter_layer_norm
_A : Optional[int] = adapter_reuse_layer_norm
_A : Any = ln_before_adapter
_A : Union[str, Any] = list(_a )
_A : List[Any] = default_language
class lowercase ( UpperCamelCase__ ):
@property
def a__ ( self ) -> Mapping[str, Mapping[int, str]]:
if self.task == "multiple-choice":
_A : Dict = {0: """batch""", 1: """choice""", 2: """sequence"""}
else:
_A : Dict = {0: """batch""", 1: """sequence"""}
return OrderedDict(
[
("""input_ids""", dynamic_axis),
("""attention_mask""", dynamic_axis),
] )
| 26 | 1 |
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_tf_available, is_torch_available
_snake_case = {
"configuration_ctrl": ["CTRL_PRETRAINED_CONFIG_ARCHIVE_MAP", "CTRLConfig"],
"tokenization_ctrl": ["CTRLTokenizer"],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"CTRL_PRETRAINED_MODEL_ARCHIVE_LIST",
"CTRLForSequenceClassification",
"CTRLLMHeadModel",
"CTRLModel",
"CTRLPreTrainedModel",
]
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"TF_CTRL_PRETRAINED_MODEL_ARCHIVE_LIST",
"TFCTRLForSequenceClassification",
"TFCTRLLMHeadModel",
"TFCTRLModel",
"TFCTRLPreTrainedModel",
]
if TYPE_CHECKING:
from .configuration_ctrl import CTRL_PRETRAINED_CONFIG_ARCHIVE_MAP, CTRLConfig
from .tokenization_ctrl import CTRLTokenizer
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_ctrl import (
CTRL_PRETRAINED_MODEL_ARCHIVE_LIST,
CTRLForSequenceClassification,
CTRLLMHeadModel,
CTRLModel,
CTRLPreTrainedModel,
)
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_tf_ctrl import (
TF_CTRL_PRETRAINED_MODEL_ARCHIVE_LIST,
TFCTRLForSequenceClassification,
TFCTRLLMHeadModel,
TFCTRLModel,
TFCTRLPreTrainedModel,
)
else:
import sys
_snake_case = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
if n == 0:
return 0
_A : Tuple = float("""-inf""" )
for i in range(1,n + 1 ):
_A : str = max(
snake_case_,prices[i - 1] + naive_cut_rod_recursive(n - i,snake_case_ ) )
return max_revue
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
_A : Dict = [float("""-inf""" ) for _ in range(n + 1 )]
return _top_down_cut_rod_recursive(snake_case_,snake_case_,snake_case_ )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
if max_rev[n] >= 0:
return max_rev[n]
elif n == 0:
return 0
else:
_A : List[str] = float("""-inf""" )
for i in range(1,n + 1 ):
_A : Optional[Any] = max(
snake_case_,prices[i - 1] + _top_down_cut_rod_recursive(n - i,snake_case_,snake_case_ ),)
_A : Tuple = max_revenue
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
# length(max_rev) = n + 1, to accommodate for the revenue obtainable from a rod of
# length 0.
_A : List[Any] = [float("""-inf""" ) for _ in range(n + 1 )]
_A : Any = 0
for i in range(1,n + 1 ):
_A : Optional[Any] = max_rev[i]
for j in range(1,i + 1 ):
_A : int = max(snake_case_,prices[j - 1] + max_rev[i - j] )
_A : int = max_revenue_i
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if n < 0:
_A : Optional[Any] = f'''n must be greater than or equal to 0. Got n = {n}'''
raise ValueError(snake_case_ )
if n > len(snake_case_ ):
_A : Any = (
"""Each integral piece of rod must have a corresponding price. """
f'''Got n = {n} but length of prices = {len(snake_case_ )}'''
)
raise ValueError(snake_case_ )
def lowerCAmelCase_ ( ):
_A : Tuple = [6, 10, 12, 15, 20, 23]
_A : List[Any] = len(snake_case_ )
# the best revenue comes from cutting the rod into 6 pieces, each
# of length 1 resulting in a revenue of 6 * 6 = 36.
_A : Any = 36
_A : List[Any] = top_down_cut_rod(snake_case_,snake_case_ )
_A : List[Any] = bottom_up_cut_rod(snake_case_,snake_case_ )
_A : Dict = naive_cut_rod_recursive(snake_case_,snake_case_ )
assert expected_max_revenue == max_rev_top_down
assert max_rev_top_down == max_rev_bottom_up
assert max_rev_bottom_up == max_rev_naive
if __name__ == "__main__":
main()
| 26 | 1 |
#
# This a `torch.distributed` diagnostics script that checks that all GPUs in the cluster (one or
# many nodes) can talk to each other via nccl and allocate gpu memory.
#
# To run first adjust the number of processes and nodes:
#
# python -m torch.distributed.run --nproc_per_node 2 --nnodes 1 torch-distributed-gpu-test.py
#
# You may need to add --master_addr $MASTER_ADDR --master_port $MASTER_PORT if using a custom addr:port
#
# You can also use the rdzv API: --rdzv_endpoint $MASTER_ADDR:$MASTER_PORT --rdzv_backend c10d
#
# use torch.distributed.launch instead of torch.distributed.run for torch < 1.9
#
# If you get a hanging in `barrier` calls you have some network issues, you may try to debug this with:
#
# NCCL_DEBUG=INFO python -m torch.distributed.run --nproc_per_node 2 --nnodes 1 torch-distributed-gpu-test.py
#
# which should tell you what's going on behind the scenes.
#
#
# This script can be run via `srun` in the SLURM environment as well. Here is a SLURM script that
# runs on 2 nodes of 4 gpus per node:
#
# #SBATCH --job-name=test-nodes # name
# #SBATCH --nodes=2 # nodes
# #SBATCH --ntasks-per-node=1 # crucial - only 1 task per dist per node!
# #SBATCH --cpus-per-task=10 # number of cores per tasks
# #SBATCH --gres=gpu:4 # number of gpus
# #SBATCH --time 0:05:00 # maximum execution time (HH:MM:SS)
# #SBATCH --output=%x-%j.out # output file name
#
# GPUS_PER_NODE=4
# MASTER_ADDR=$(scontrol show hostnames $SLURM_JOB_NODELIST | head -n 1)
# MASTER_PORT=6000
#
# srun --jobid $SLURM_JOBID bash -c 'python -m torch.distributed.run \
# --nproc_per_node $GPUS_PER_NODE --nnodes $SLURM_NNODES --node_rank $SLURM_PROCID \
# --master_addr $MASTER_ADDR --master_port $MASTER_PORT \
# torch-distributed-gpu-test.py'
#
import fcntl
import os
import socket
import torch
import torch.distributed as dist
def lowerCAmelCase_ ( *snake_case_ ):
with open(snake_case_,"""r""" ) as fh:
fcntl.flock(snake_case_,fcntl.LOCK_EX )
try:
print(*snake_case_ )
finally:
fcntl.flock(snake_case_,fcntl.LOCK_UN )
_snake_case = int(os.environ["LOCAL_RANK"])
torch.cuda.set_device(local_rank)
_snake_case = torch.device("cuda", local_rank)
_snake_case = socket.gethostname()
_snake_case = f"""[{hostname}-{local_rank}]"""
try:
# test distributed
dist.init_process_group("nccl")
dist.all_reduce(torch.ones(1).to(device), op=dist.ReduceOp.SUM)
dist.barrier()
# test cuda is available and can allocate memory
torch.cuda.is_available()
torch.ones(1).cuda(local_rank)
# global rank
_snake_case = dist.get_rank()
_snake_case = dist.get_world_size()
printflock(f"""{gpu} is OK (global rank: {rank}/{world_size})""")
dist.barrier()
if rank == 0:
printflock(f"""pt={torch.__version__}, cuda={torch.version.cuda}, nccl={torch.cuda.nccl.version()}""")
except Exception:
printflock(f"""{gpu} is broken""")
raise
| 26 |
import requests
from bsa import BeautifulSoup
def lowerCAmelCase_ ( snake_case_ = "AAPL" ):
_A : str = f'''https://in.finance.yahoo.com/quote/{symbol}?s={symbol}'''
_A : List[Any] = BeautifulSoup(requests.get(snake_case_ ).text,"""html.parser""" )
_A : Union[str, Any] = """My(6px) Pos(r) smartphone_Mt(6px)"""
return soup.find("""div""",class_=class_ ).find("""span""" ).text
if __name__ == "__main__":
for symbol in "AAPL AMZN IBM GOOG MSFT ORCL".split():
print(f"""Current {symbol:<4} stock price is {stock_price(symbol):>8}""")
| 26 | 1 |
from math import ceil, sqrt
def lowerCAmelCase_ ( snake_case_ = 1000000 ):
_A : int = 0
for outer_width in range(3,(limit // 4) + 2 ):
if outer_width**2 > limit:
_A : List[str] = max(ceil(sqrt(outer_width**2 - limit ) ),1 )
else:
_A : Optional[int] = 1
if (outer_width - hole_width_lower_bound) % 2:
hole_width_lower_bound += 1
answer += (outer_width - hole_width_lower_bound - 2) // 2 + 1
return answer
if __name__ == "__main__":
print(f"""{solution() = }""")
| 26 |
import unittest
from transformers import (
MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
TextaTextGenerationPipeline,
pipeline,
)
from transformers.testing_utils import is_pipeline_test, require_tf, require_torch
from transformers.utils import is_torch_available
from .test_pipelines_common import ANY
if is_torch_available():
import torch
@is_pipeline_test
class lowercase ( unittest.TestCase ):
_a = MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING
_a = TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING
def a__ ( self , _a , _a , _a ) -> int:
_A : str = TextaTextGenerationPipeline(model=_a , tokenizer=_a )
return generator, ["Something to write", "Something else"]
def a__ ( self , _a , _a ) -> Dict:
_A : Any = generator("""Something there""" )
self.assertEqual(_a , [{"""generated_text""": ANY(_a )}] )
# These are encoder decoder, they don't just append to incoming string
self.assertFalse(outputs[0]["""generated_text"""].startswith("""Something there""" ) )
_A : List[Any] = generator(["""This is great !""", """Something else"""] , num_return_sequences=2 , do_sample=_a )
self.assertEqual(
_a , [
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
] , )
_A : Optional[int] = generator(
["""This is great !""", """Something else"""] , num_return_sequences=2 , batch_size=2 , do_sample=_a )
self.assertEqual(
_a , [
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
] , )
with self.assertRaises(_a ):
generator(4 )
@require_torch
def a__ ( self ) -> List[str]:
_A : Any = pipeline("""text2text-generation""" , model="""patrickvonplaten/t5-tiny-random""" , framework="""pt""" )
# do_sample=False necessary for reproducibility
_A : Dict = generator("""Something there""" , do_sample=_a )
self.assertEqual(_a , [{"""generated_text""": """"""}] )
_A : Any = 3
_A : Any = generator(
"""Something there""" , num_return_sequences=_a , num_beams=_a , )
_A : Optional[int] = [
{"""generated_text""": """Beide Beide Beide Beide Beide Beide Beide Beide Beide"""},
{"""generated_text""": """Beide Beide Beide Beide Beide Beide Beide Beide"""},
{"""generated_text""": """"""},
]
self.assertEqual(_a , _a )
_A : Dict = generator("""This is a test""" , do_sample=_a , num_return_sequences=2 , return_tensors=_a )
self.assertEqual(
_a , [
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
] , )
_A : Dict = generator.model.config.eos_token_id
_A : List[str] = """<pad>"""
_A : Dict = generator(
["""This is a test""", """This is a second test"""] , do_sample=_a , num_return_sequences=2 , batch_size=2 , return_tensors=_a , )
self.assertEqual(
_a , [
[
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
],
[
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
],
] , )
@require_tf
def a__ ( self ) -> int:
_A : Optional[Any] = pipeline("""text2text-generation""" , model="""patrickvonplaten/t5-tiny-random""" , framework="""tf""" )
# do_sample=False necessary for reproducibility
_A : str = generator("""Something there""" , do_sample=_a )
self.assertEqual(_a , [{"""generated_text""": """"""}] )
| 26 | 1 |
import os
import unicodedata
from shutil import copyfile
from typing import Any, Dict, List, Optional, Tuple
import sentencepiece as spm
from ...tokenization_utils import AddedToken, PreTrainedTokenizer
from ...utils import SPIECE_UNDERLINE, logging
_snake_case = logging.get_logger(__name__)
_snake_case = {"vocab_file": "spiece.model"}
_snake_case = {
"vocab_file": {
"TsinghuaAI/CPM-Generate": "https://huggingface.co/TsinghuaAI/CPM-Generate/resolve/main/spiece.model",
}
}
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a=False , _a=True , _a=False , _a="<s>" , _a="</s>" , _a="<unk>" , _a="<sep>" , _a="<pad>" , _a="<cls>" , _a="<mask>" , _a=["<eop>", "<eod>"] , _a = None , **_a , ) -> None:
_A : Dict = AddedToken(_a , lstrip=_a , rstrip=_a ) if isinstance(_a , _a ) else mask_token
_A : int = {} if sp_model_kwargs is None else sp_model_kwargs
super().__init__(
do_lower_case=_a , remove_space=_a , keep_accents=_a , bos_token=_a , eos_token=_a , unk_token=_a , sep_token=_a , pad_token=_a , cls_token=_a , mask_token=_a , additional_special_tokens=_a , sp_model_kwargs=self.sp_model_kwargs , **_a , )
_A : List[str] = 3
_A : List[str] = do_lower_case
_A : Dict = remove_space
_A : Optional[int] = keep_accents
_A : Union[str, Any] = vocab_file
_A : int = spm.SentencePieceProcessor(**self.sp_model_kwargs )
self.sp_model.Load(_a )
try:
import jieba
except ModuleNotFoundError as error:
raise error.__class__(
"""You need to install jieba to use CpmTokenizer or CpmTokenizerFast. """
"""See https://pypi.org/project/jieba/ for installation.""" )
_A : Optional[int] = jieba
_A : Union[str, Any] = str.maketrans(""" \n""" , """\u2582\u2583""" )
@property
# Copied from transformers.models.xlnet.tokenization_xlnet.XLNetTokenizer.vocab_size
def a__ ( self ) -> List[Any]:
return len(self.sp_model )
def a__ ( self ) -> Tuple:
_A : Optional[Any] = {self.convert_ids_to_tokens(_a ): i for i in range(self.vocab_size )}
vocab.update(self.added_tokens_encoder )
return vocab
def __getstate__( self ) -> Tuple:
_A : List[Any] = self.__dict__.copy()
_A : Optional[int] = None
return state
def __setstate__( self , _a ) -> Union[str, Any]:
_A : Dict = d
# for backward compatibility
if not hasattr(self , """sp_model_kwargs""" ):
_A : Optional[int] = {}
_A : int = spm.SentencePieceProcessor(**self.sp_model_kwargs )
self.sp_model.Load(self.vocab_file )
def a__ ( self , _a ) -> str:
if self.remove_space:
_A : Dict = """ """.join(inputs.strip().split() )
else:
_A : List[str] = inputs
_A : str = outputs.replace("""``""" , """\"""" ).replace("""''""" , """\"""" )
if not self.keep_accents:
_A : List[str] = unicodedata.normalize("""NFKD""" , _a )
_A : Dict = """""".join([c for c in outputs if not unicodedata.combining(_a )] )
if self.do_lower_case:
_A : Union[str, Any] = outputs.lower()
return outputs
def a__ ( self , _a ) -> List[str]:
_A : List[str] = self.preprocess_text(_a )
_A : Tuple = self.sp_model.encode(_a , out_type=_a )
_A : List[str] = []
for piece in pieces:
if len(_a ) > 1 and piece[-1] == str(""",""" ) and piece[-2].isdigit():
_A : List[Any] = self.sp_model.EncodeAsPieces(piece[:-1].replace(_a , """""" ) )
if piece[0] != SPIECE_UNDERLINE and cur_pieces[0][0] == SPIECE_UNDERLINE:
if len(cur_pieces[0] ) == 1:
_A : int = cur_pieces[1:]
else:
_A : str = cur_pieces[0][1:]
cur_pieces.append(piece[-1] )
new_pieces.extend(_a )
else:
new_pieces.append(_a )
return new_pieces
def a__ ( self , _a ) -> Union[str, Any]:
return self.sp_model.PieceToId(_a )
def a__ ( self , _a ) -> Optional[Any]:
return self.sp_model.IdToPiece(_a )
def a__ ( self , _a ) -> int:
_A : Tuple = """""".join(_a ).replace(_a , """ """ ).strip()
return out_string
def a__ ( self , _a , _a = None ) -> List[int]:
_A : Dict = [self.sep_token_id]
_A : Union[str, Any] = [self.cls_token_id]
if token_ids_a is None:
return token_ids_a + sep + cls
return token_ids_a + sep + token_ids_a + sep + cls
def a__ ( self , _a , _a = None , _a = False ) -> List[int]:
if already_has_special_tokens:
return super().get_special_tokens_mask(
token_ids_a=_a , token_ids_a=_a , already_has_special_tokens=_a )
if token_ids_a is not None:
return ([0] * len(_a )) + [1] + ([0] * len(_a )) + [1, 1]
return ([0] * len(_a )) + [1, 1]
def a__ ( self , _a , _a = None ) -> List[int]:
_A : Union[str, Any] = [self.sep_token_id]
_A : Any = [2]
if token_ids_a is None:
return len(token_ids_a + sep ) * [0] + cls_segment_id
return len(token_ids_a + sep ) * [0] + len(token_ids_a + sep ) * [1] + cls_segment_id
def a__ ( self , _a , _a = None ) -> Tuple[str]:
if not os.path.isdir(_a ):
logger.error(F'''Vocabulary path ({save_directory}) should be a directory''' )
return
_A : Optional[int] = os.path.join(
_a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
if os.path.abspath(self.vocab_file ) != os.path.abspath(_a ) and os.path.isfile(self.vocab_file ):
copyfile(self.vocab_file , _a )
elif not os.path.isfile(self.vocab_file ):
with open(_a , """wb""" ) as fi:
_A : List[Any] = self.sp_model.serialized_model_proto()
fi.write(_a )
return (out_vocab_file,)
def a__ ( self , *_a , **_a ) -> Optional[Any]:
_A : Optional[Any] = super()._decode(*_a , **_a )
_A : Optional[int] = text.replace(""" """ , """""" ).replace("""\u2582""" , """ """ ).replace("""\u2583""" , """\n""" )
return text
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
while b:
_A , _A : List[str] = b, a % b
return a
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
return a if b == 0 else euclidean_gcd_recursive(snake_case_,a % b )
def lowerCAmelCase_ ( ):
print(f'''euclidean_gcd(3, 5) = {euclidean_gcd(3,5 )}''' )
print(f'''euclidean_gcd(5, 3) = {euclidean_gcd(5,3 )}''' )
print(f'''euclidean_gcd(1, 3) = {euclidean_gcd(1,3 )}''' )
print(f'''euclidean_gcd(3, 6) = {euclidean_gcd(3,6 )}''' )
print(f'''euclidean_gcd(6, 3) = {euclidean_gcd(6,3 )}''' )
print(f'''euclidean_gcd_recursive(3, 5) = {euclidean_gcd_recursive(3,5 )}''' )
print(f'''euclidean_gcd_recursive(5, 3) = {euclidean_gcd_recursive(5,3 )}''' )
print(f'''euclidean_gcd_recursive(1, 3) = {euclidean_gcd_recursive(1,3 )}''' )
print(f'''euclidean_gcd_recursive(3, 6) = {euclidean_gcd_recursive(3,6 )}''' )
print(f'''euclidean_gcd_recursive(6, 3) = {euclidean_gcd_recursive(6,3 )}''' )
if __name__ == "__main__":
main()
| 26 | 1 |
import collections
import os
from typing import List, Optional, Tuple
from transformers.utils import is_jieba_available, requires_backends
if is_jieba_available():
import jieba
from ...tokenization_utils import PreTrainedTokenizer
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {"vocab_file": "vocab.txt"}
_snake_case = {
"vocab_file": {
"openbmb/cpm-ant-10b": "https://huggingface.co/openbmb/cpm-ant-10b/blob/main/vocab.txt",
},
}
_snake_case = {
"openbmb/cpm-ant-10b": 1024,
}
def lowerCAmelCase_ ( snake_case_ ):
_A : Tuple = collections.OrderedDict()
with open(snake_case_,"""r""",encoding="""utf-8""" ) as reader:
_A : str = reader.readlines()
for index, token in enumerate(snake_case_ ):
_A : Any = token.rstrip("""\n""" )
_A : List[str] = index
return vocab
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a="<unk>" , _a=200 ) -> str:
_A : Optional[Any] = vocab
_A : str = unk_token
_A : Optional[int] = max_input_chars_per_word
def a__ ( self , _a ) -> int:
_A : Optional[int] = list(_a )
if len(_a ) > self.max_input_chars_per_word:
return [self.unk_token]
_A : Union[str, Any] = 0
_A : List[str] = []
while start < len(_a ):
_A : int = len(_a )
_A : Any = None
while start < end:
_A : Optional[int] = """""".join(chars[start:end] )
if substr in self.vocab:
_A : Dict = substr
break
end -= 1
if cur_substr is None:
sub_tokens.append(self.unk_token )
start += 1
else:
sub_tokens.append(_a )
_A : Any = end
return sub_tokens
class lowercase ( UpperCamelCase__ ):
_a = VOCAB_FILES_NAMES
_a = PRETRAINED_VOCAB_FILES_MAP
_a = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
_a = ["input_ids", "attention_mask"]
_a = False
def __init__( self , _a , _a="<d>" , _a="</d>" , _a="<s>" , _a="</s>" , _a="<pad>" , _a="<unk>" , _a="</n>" , _a="</_>" , _a="left" , **_a , ) -> Tuple:
requires_backends(self , ["""jieba"""] )
super().__init__(
bod_token=_a , eod_token=_a , bos_token=_a , eos_token=_a , pad_token=_a , unk_token=_a , line_token=_a , space_token=_a , padding_side=_a , **_a , )
_A : List[Any] = bod_token
_A : List[str] = eod_token
_A : List[str] = load_vocab(_a )
_A : List[str] = self.encoder[space_token]
_A : Tuple = self.encoder[line_token]
del self.encoder[space_token]
del self.encoder[line_token]
_A : Optional[int] = collections.OrderedDict(sorted(self.encoder.items() , key=lambda _a : x[1] ) )
_A : str = {v: k for k, v in self.encoder.items()}
_A : List[Any] = WordpieceTokenizer(vocab=self.encoder , unk_token=self.unk_token )
@property
def a__ ( self ) -> str:
return self.encoder[self.bod_token]
@property
def a__ ( self ) -> int:
return self.encoder[self.eod_token]
@property
def a__ ( self ) -> Optional[Any]:
return self.encoder["\n"]
@property
def a__ ( self ) -> int:
return len(self.encoder )
def a__ ( self ) -> List[Any]:
return dict(self.encoder , **self.added_tokens_encoder )
def a__ ( self , _a ) -> Dict:
_A : Optional[int] = []
for x in jieba.cut(_a , cut_all=_a ):
output_tokens.extend(self.wordpiece_tokenizer.tokenize(_a ) )
return output_tokens
def a__ ( self , _a , **_a ) -> str:
_A : List[str] = [i for i in token_ids if i >= 0]
_A : Dict = [
x for x in token_ids if x != self.pad_token_id and x != self.eos_token_id and x != self.bos_token_id
]
return super()._decode(_a , **_a )
def a__ ( self , _a ) -> Union[str, Any]:
return token in self.encoder
def a__ ( self , _a ) -> str:
return "".join(_a )
def a__ ( self , _a ) -> str:
return self.encoder.get(_a , self.encoder.get(self.unk_token ) )
def a__ ( self , _a ) -> Tuple:
return self.decoder.get(_a , self.unk_token )
def a__ ( self , _a , _a = None ) -> Tuple[str]:
if os.path.isdir(_a ):
_A : List[Any] = os.path.join(
_a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
else:
_A : Optional[int] = (filename_prefix + """-""" if filename_prefix else """""") + save_directory
_A : Dict = 0
if " " in self.encoder:
_A : str = self.encoder[""" """]
del self.encoder[" "]
if "\n" in self.encoder:
_A : Optional[int] = self.encoder["""\n"""]
del self.encoder["\n"]
_A : int = collections.OrderedDict(sorted(self.encoder.items() , key=lambda _a : x[1] ) )
with open(_a , """w""" , encoding="""utf-8""" ) as writer:
for token, token_index in self.encoder.items():
if index != token_index:
logger.warning(
F'''Saving vocabulary to {vocab_file}: vocabulary indices are not consecutive.'''
""" Please check that the vocabulary is not corrupted!""" )
_A : str = token_index
writer.write(token + """\n""" )
index += 1
return (vocab_file,)
def a__ ( self , _a , _a = None ) -> List[int]:
if token_ids_a is None:
return [self.bos_token_id] + token_ids_a
return [self.bos_token_id] + token_ids_a + [self.bos_token_id] + token_ids_a
def a__ ( self , _a , _a = None , _a = False ) -> List[int]:
if already_has_special_tokens:
return super().get_special_tokens_mask(
token_ids_a=_a , token_ids_a=_a , already_has_special_tokens=_a )
if token_ids_a is not None:
return [1] + ([0] * len(_a )) + [1] + ([0] * len(_a ))
return [1] + ([0] * len(_a ))
| 26 |
def lowerCAmelCase_ ( snake_case_ ):
if number < 0:
raise ValueError("""number must not be negative""" )
return number & (number - 1) == 0
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
from math import ceil
from typing import List, Optional, Union
import numpy as np
from ...audio_utils import mel_filter_bank, spectrogram, window_function
from ...feature_extraction_sequence_utils import BatchFeature, SequenceFeatureExtractor
from ...utils import TensorType, logging
_snake_case = logging.get_logger(__name__)
class lowercase ( UpperCamelCase__ ):
_a = ["audio_values", "audio_mask"]
def __init__( self , _a=2048 , _a=1 , _a=[16, 16] , _a=128 , _a=4_4100 , _a=86 , _a=2048 , _a=0.0 , **_a , ) -> List[Any]:
super().__init__(
feature_size=_a , sampling_rate=_a , padding_value=_a , **_a , )
_A : Any = spectrogram_length
_A : Dict = num_channels
_A : Optional[Any] = patch_size
_A : str = feature_size // self.patch_size[1]
_A : List[Any] = n_fft
_A : Optional[Any] = sampling_rate // hop_length_to_sampling_rate
_A : List[str] = sampling_rate
_A : Union[str, Any] = padding_value
_A : List[str] = mel_filter_bank(
num_frequency_bins=1 + n_fft // 2 , num_mel_filters=_a , min_frequency=0.0 , max_frequency=22050.0 , sampling_rate=_a , norm="""slaney""" , mel_scale="""slaney""" , ).T
def a__ ( self , _a ) -> np.ndarray:
_A : Optional[int] = spectrogram(
_a , window_function(self.n_fft , """hann""" ) , frame_length=self.n_fft , hop_length=self.hop_length , power=2.0 , mel_filters=self.mel_filters.T , log_mel="""dB""" , db_range=80.0 , )
_A : Tuple = log_spec[:, :-1]
_A : int = log_spec - 20.0
_A : Tuple = np.clip(log_spec / 40.0 , -2.0 , 0.0 ) + 1.0
return log_spec
def __call__( self , _a , _a = None , _a = True , _a = None , _a = False , _a = False , **_a , ) -> BatchFeature:
if sampling_rate is not None:
if sampling_rate != self.sampling_rate:
raise ValueError(
"""This feature extractor is set to support sampling rate"""
F''' of {self.sampling_rate}. Please make sure that the provided `raw_speech` input was sampled'''
F''' with {self.sampling_rate} and not {sampling_rate}.''' )
else:
logger.warning(
"""It is strongly recommended to pass the `sampling_rate` argument to this function. """
"""Failing to do so can result in silent errors that might be hard to debug.""" )
_A : int = isinstance(_a , np.ndarray ) and len(raw_speech.shape ) > 1
if is_batched_numpy and len(raw_speech.shape ) > 2:
raise ValueError(F'''Only mono-channel audio is supported for input to {self}''' )
_A : Optional[Any] = is_batched_numpy or (
isinstance(_a , (list, tuple) ) and (isinstance(raw_speech[0] , (np.ndarray, tuple, list) ))
)
if is_batched:
_A : Union[str, Any] = [np.asarray([speech] , dtype=np.floataa ).T for speech in raw_speech]
elif not is_batched and not isinstance(_a , np.ndarray ):
_A : str = np.asarray(_a , dtype=np.floataa )
elif isinstance(_a , np.ndarray ) and raw_speech.dtype is np.dtype(np.floataa ):
_A : Dict = raw_speech.astype(np.floataa )
# always return batch
if not is_batched:
_A : Dict = [np.asarray([raw_speech] ).T]
# Convert audio signals to log mel spectrograms, truncate by time axis
_A : Union[str, Any] = [
self._np_extract_fbank_features(waveform.squeeze() ).T[: self.spectrogram_length] for waveform in raw_speech
]
if isinstance(audio_features[0] , _a ):
_A : Optional[int] = [np.asarray(_a , dtype=np.floataa ) for feature in audio_features]
# Create audio attention mask
_A : Any = max(
[ceil(feature.shape[0] / self.patch_size[0] ) * self.freq_len for feature in audio_features] ) # The maximum number of audio patches in a batch
if return_attention_mask:
_A : str = [
(ceil(feature.shape[0] / self.patch_size[0] ) * self.freq_len) * [1]
+ (max_patch_len - ceil(feature.shape[0] / self.patch_size[0] ) * self.freq_len) * [0]
for feature in audio_features
]
_A : List[str] = np.array(_a ).astype(np.floataa )
# convert into correct format for padding
_A : Tuple = max_patch_len // self.freq_len * self.patch_size[0] # The maximum audio size in a batch
_A : int = np.ones([len(_a ), 1, max_time_len, self.feature_size] ).astype(np.floataa )
_A : Optional[int] = padded_audio_features * self.padding_value
for i in range(len(_a ) ):
_A : Union[str, Any] = audio_features[i]
_A : Optional[Any] = feature
# return as BatchFeature
if return_attention_mask:
_A : Optional[Any] = {"""audio_values""": padded_audio_features, """audio_mask""": audio_mask}
else:
_A : Any = {"""audio_values""": padded_audio_features}
_A : int = BatchFeature(data=_a , tensor_type=_a )
return encoded_inputs
| 26 |
import argparse
import torch
from transformers import BlenderbotConfig, BlenderbotForConditionalGeneration
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
_snake_case = [
["attention", "attn"],
["encoder_attention", "encoder_attn"],
["q_lin", "q_proj"],
["k_lin", "k_proj"],
["v_lin", "v_proj"],
["out_lin", "out_proj"],
["norm_embeddings", "layernorm_embedding"],
["position_embeddings", "embed_positions"],
["embeddings", "embed_tokens"],
["ffn.lin", "fc"],
]
def lowerCAmelCase_ ( snake_case_ ):
if k == "embeddings.weight":
return "shared.weight"
for parlai_name, hf_name in PATTERNS:
_A : str = k.replace(snake_case_,snake_case_ )
if k.startswith("""encoder""" ):
_A : Optional[Any] = k.replace(""".attn""",""".self_attn""" )
_A : Dict = k.replace("""norm1""","""self_attn_layer_norm""" )
_A : Optional[Any] = k.replace("""norm2""","""final_layer_norm""" )
elif k.startswith("""decoder""" ):
_A : str = k.replace("""norm1""","""self_attn_layer_norm""" )
_A : Any = k.replace("""norm2""","""encoder_attn_layer_norm""" )
_A : Optional[int] = k.replace("""norm3""","""final_layer_norm""" )
return k
def lowerCAmelCase_ ( snake_case_ ):
_A : List[Any] = [
"""model.encoder.layernorm_embedding.weight""",
"""model.encoder.layernorm_embedding.bias""",
"""model.decoder.layernorm_embedding.weight""",
"""model.decoder.layernorm_embedding.bias""",
]
for k in keys:
_A : str = sd.pop(snake_case_ )
_A : Optional[int] = k.replace("""layernorm_embedding""","""layer_norm""" )
assert new_k not in sd
_A : Optional[int] = v
_snake_case = ["START"]
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
_A : List[Any] = model["""model"""]
_A : Optional[Any] = BlenderbotConfig.from_json_file(snake_case_ )
_A : List[str] = BlenderbotForConditionalGeneration(snake_case_ )
_A : Tuple = m.model.state_dict().keys()
_A : Any = []
_A : Dict = {}
for k, v in sd.items():
if k in IGNORE_KEYS:
continue
_A : Optional[int] = rename_state_dict_key(snake_case_ )
if new_k not in valid_keys:
failures.append([k, new_k] )
else:
_A : Dict = v
if cfg.normalize_before: # Blenderbot-3B checkpoints. Rename layernorm_embedding -> layer_norm
rename_layernorm_keys(snake_case_ )
m.model.load_state_dict(snake_case_,strict=snake_case_ )
m.half()
m.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument("--src_path", type=str, help="like blenderbot-model.bin")
parser.add_argument("--save_dir", default="hf_blenderbot", type=str, help="Where to save converted model.")
parser.add_argument(
"--hf_config_json", default="blenderbot-3b-config.json", type=str, help="Path to config to use"
)
_snake_case = parser.parse_args()
convert_parlai_checkpoint(args.src_path, args.save_dir, args.hf_config_json)
| 26 | 1 |
import random
import timeit
from functools import wraps
from typing import Callable, Optional
from ..configuration_utils import PretrainedConfig
from ..models.auto.modeling_tf_auto import TF_MODEL_MAPPING, TF_MODEL_WITH_LM_HEAD_MAPPING
from ..utils import is_pyanvml_available, is_tf_available, logging
from .benchmark_utils import (
Benchmark,
Memory,
MemorySummary,
measure_peak_memory_cpu,
start_memory_tracing,
stop_memory_tracing,
)
if is_tf_available():
import tensorflow as tf
from tensorflow.python.framework.errors_impl import ResourceExhaustedError
from .benchmark_args_tf import TensorFlowBenchmarkArguments
if is_pyanvml_available():
import pyanvml.pyanvml as nvml
_snake_case = logging.get_logger(__name__)
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
def run_func(snake_case_ ):
@wraps(snake_case_ )
def run_in_eager_mode(*snake_case_,**snake_case_ ):
return func(*snake_case_,**snake_case_ )
@wraps(snake_case_ )
@tf.function(experimental_compile=snake_case_ )
def run_in_graph_mode(*snake_case_,**snake_case_ ):
return func(*snake_case_,**snake_case_ )
if do_eager_mode is True:
if use_xla is not False:
raise ValueError(
"""Cannot run model in XLA, if `args.eager_mode` is set to `True`. Please set `args.eager_mode=False`.""" )
return run_in_eager_mode
else:
return run_in_graph_mode
return run_func
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : List[Any] = random.Random()
_A : Optional[int] = [rng.randint(0,vocab_size - 1 ) for i in range(batch_size * sequence_length )]
return tf.constant(snake_case_,shape=(batch_size, sequence_length),dtype=tf.intaa )
class lowercase ( UpperCamelCase__ ):
_a = 42
_a = 42
_a = "TensorFlow"
@property
def a__ ( self ) -> Any:
return tf.__version__
def a__ ( self , _a , _a , _a ) -> float:
# initialize GPU on separate process
_A : List[Any] = self.args.strategy
if strategy is None:
raise ValueError("""A device strategy has to be initialized before using TensorFlow.""" )
_A : Optional[Any] = self._prepare_inference_func(_a , _a , _a )
return self._measure_speed(_inference )
def a__ ( self , _a , _a , _a ) -> float:
_A : Optional[int] = self.args.strategy
if strategy is None:
raise ValueError("""A device strategy has to be initialized before using TensorFlow.""" )
_A : Dict = self._prepare_train_func(_a , _a , _a )
return self._measure_speed(_train )
def a__ ( self , _a , _a , _a ) -> [Memory, Optional[MemorySummary]]:
# initialize GPU on separate process
if self.args.is_gpu:
tf.config.experimental.set_memory_growth(self.args.gpu_list[self.args.device_idx] , _a )
_A : Dict = self.args.strategy
if strategy is None:
raise ValueError("""A device strategy has to be initialized before using TensorFlow.""" )
_A : Dict = self._prepare_inference_func(_a , _a , _a )
return self._measure_memory(_inference )
def a__ ( self , _a , _a , _a ) -> [Memory, Optional[MemorySummary]]:
if self.args.is_gpu:
tf.config.experimental.set_memory_growth(self.args.gpu_list[self.args.device_idx] , _a )
_A : str = self.args.strategy
if strategy is None:
raise ValueError("""A device strategy has to be initialized before using TensorFlow.""" )
_A : str = self._prepare_train_func(_a , _a , _a )
return self._measure_memory(_train )
def a__ ( self , _a , _a , _a ) -> Callable[[], None]:
_A : Optional[Any] = self.config_dict[model_name]
if self.args.fpaa:
raise NotImplementedError("""Mixed precision is currently not supported.""" )
_A : List[Any] = (
hasattr(_a , """architectures""" )
and isinstance(config.architectures , _a )
and len(config.architectures ) > 0
)
if not self.args.only_pretrain_model and has_model_class_in_config:
try:
_A : Dict = """TF""" + config.architectures[0] # prepend 'TF' for tensorflow model
_A : Tuple = __import__("""transformers""" , fromlist=[model_class] )
_A : Tuple = getattr(_a , _a )
_A : Optional[int] = model_cls(_a )
except ImportError:
raise ImportError(
F'''{model_class} does not exist. If you just want to test the pretrained model, you might want to'''
""" set `--only_pretrain_model` or `args.only_pretrain_model=True`.""" )
else:
_A : Union[str, Any] = TF_MODEL_MAPPING[config.__class__](_a )
# encoder-decoder has vocab size saved differently
_A : Any = config.vocab_size if hasattr(_a , """vocab_size""" ) else config.encoder.vocab_size
_A : Dict = random_input_ids(_a , _a , _a )
@run_with_tf_optimizations(self.args.eager_mode , self.args.use_xla )
def encoder_decoder_forward():
return model(_a , decoder_input_ids=_a , training=_a )
@run_with_tf_optimizations(self.args.eager_mode , self.args.use_xla )
def encoder_forward():
return model(_a , training=_a )
_A : str = encoder_decoder_forward if config.is_encoder_decoder else encoder_forward
return _inference
def a__ ( self , _a , _a , _a ) -> Callable[[], None]:
_A : str = self.config_dict[model_name]
if self.args.eager_mode is not False:
raise ValueError("""Training cannot be done in eager mode. Please make sure that `args.eager_mode = False`.""" )
if self.args.fpaa:
raise NotImplementedError("""Mixed precision is currently not supported.""" )
_A : str = (
hasattr(_a , """architectures""" )
and isinstance(config.architectures , _a )
and len(config.architectures ) > 0
)
if not self.args.only_pretrain_model and has_model_class_in_config:
try:
_A : List[str] = """TF""" + config.architectures[0] # prepend 'TF' for tensorflow model
_A : Union[str, Any] = __import__("""transformers""" , fromlist=[model_class] )
_A : Any = getattr(_a , _a )
_A : List[Any] = model_cls(_a )
except ImportError:
raise ImportError(
F'''{model_class} does not exist. If you just want to test the pretrained model, you might want to'''
""" set `--only_pretrain_model` or `args.only_pretrain_model=True`.""" )
else:
_A : Optional[Any] = TF_MODEL_WITH_LM_HEAD_MAPPING[config.__class__](_a )
# encoder-decoder has vocab size saved differently
_A : str = config.vocab_size if hasattr(_a , """vocab_size""" ) else config.encoder.vocab_size
_A : Optional[int] = random_input_ids(_a , _a , _a )
@run_with_tf_optimizations(self.args.eager_mode , self.args.use_xla )
def encoder_decoder_train():
_A : str = model(_a , decoder_input_ids=_a , labels=_a , training=_a )[0]
_A : List[str] = tf.gradients(_a , model.trainable_variables )
return gradients
@run_with_tf_optimizations(self.args.eager_mode , self.args.use_xla )
def encoder_train():
_A : str = model(_a , labels=_a , training=_a )[0]
_A : Tuple = tf.gradients(_a , model.trainable_variables )
return gradients
_A : List[str] = encoder_decoder_train if config.is_encoder_decoder else encoder_train
return _train
def a__ ( self , _a ) -> float:
with self.args.strategy.scope():
try:
if self.args.is_tpu or self.args.use_xla:
# run additional 10 times to stabilize compilation for tpu
logger.info("""Do inference on TPU. Running model 5 times to stabilize compilation""" )
timeit.repeat(_a , repeat=1 , number=5 )
# as written in https://docs.python.org/2/library/timeit.html#timeit.Timer.repeat, min should be taken rather than the average
_A : Union[str, Any] = timeit.repeat(
_a , repeat=self.args.repeat , number=10 , )
return min(_a ) / 10.0
except ResourceExhaustedError as e:
self.print_fn(F'''Doesn\'t fit on GPU. {e}''' )
def a__ ( self , _a ) -> [Memory, MemorySummary]:
logger.info(
"""Note that TensorFlow allocates more memory than """
"""it might need to speed up computation. """
"""The memory reported here corresponds to the memory """
"""reported by `nvidia-smi`, which can vary depending """
"""on total available memory on the GPU that is used.""" )
with self.args.strategy.scope():
try:
if self.args.trace_memory_line_by_line:
if not self.args.eager_mode:
raise ValueError(
"""`args.eager_mode` is set to `False`. Make sure to run model in eager mode to measure memory"""
""" consumption line by line.""" )
_A : int = start_memory_tracing("""transformers""" )
if self.args.is_tpu:
# tpu
raise NotImplementedError(
"""Memory Benchmarking is currently not implemented for TPU. Please disable memory benchmarking"""
""" with `args.memory=False`""" )
elif self.args.is_gpu:
# gpu
if not is_pyanvml_available():
logger.warning(
"""py3nvml not installed, we won't log GPU memory usage. """
"""Install py3nvml (pip install py3nvml) to log information about GPU.""" )
_A : List[str] = """N/A"""
else:
logger.info(
"""Measuring total GPU usage on GPU device. Make sure to not have additional processes"""
""" running on the same GPU.""" )
# init nvml
nvml.nvmlInit()
func()
_A : str = nvml.nvmlDeviceGetHandleByIndex(self.args.device_idx )
_A : Optional[int] = nvml.nvmlDeviceGetMemoryInfo(_a )
_A : int = meminfo.used
_A : Tuple = Memory(_a )
# shutdown nvml
nvml.nvmlShutdown()
else:
# cpu
if self.args.trace_memory_line_by_line:
logger.info(
"""When enabling line by line tracing, the max peak memory for CPU is inaccurate in"""
""" TensorFlow.""" )
_A : Optional[Any] = None
else:
_A : Tuple = measure_peak_memory_cpu(_a )
_A : int = Memory(_a ) if isinstance(_a , _a ) else memory_bytes
if self.args.trace_memory_line_by_line:
_A : Optional[int] = stop_memory_tracing(_a )
if memory is None:
_A : Tuple = summary.total
else:
_A : Optional[Any] = None
return memory, summary
except ResourceExhaustedError as e:
self.print_fn(F'''Doesn\'t fit on GPU. {e}''' )
return "N/A", None
| 26 |
import multiprocessing
from typing import TYPE_CHECKING, Optional, Union
from .. import Dataset, Features, config
from ..formatting import query_table
from ..packaged_modules.sql.sql import Sql
from ..utils import logging
from .abc import AbstractDatasetInputStream
if TYPE_CHECKING:
import sqlitea
import sqlalchemy
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a = None , _a = None , _a = False , **_a , ) -> int:
super().__init__(features=_a , cache_dir=_a , keep_in_memory=_a , **_a )
_A : Optional[int] = Sql(
cache_dir=_a , features=_a , sql=_a , con=_a , **_a , )
def a__ ( self ) -> Optional[Any]:
_A : Tuple = None
_A : int = None
_A : Tuple = None
_A : Union[str, Any] = None
self.builder.download_and_prepare(
download_config=_a , download_mode=_a , verification_mode=_a , base_path=_a , )
# Build dataset for splits
_A : int = self.builder.as_dataset(
split="""train""" , verification_mode=_a , in_memory=self.keep_in_memory )
return dataset
class lowercase :
def __init__( self , _a , _a , _a , _a = None , _a = None , **_a , ) -> Union[str, Any]:
if num_proc is not None and num_proc <= 0:
raise ValueError(F'''num_proc {num_proc} must be an integer > 0.''' )
_A : Dict = dataset
_A : int = name
_A : Union[str, Any] = con
_A : str = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE
_A : str = num_proc
_A : Optional[Any] = to_sql_kwargs
def a__ ( self ) -> int:
_A : Any = self.to_sql_kwargs.pop("""sql""" , _a )
_A : List[str] = self.to_sql_kwargs.pop("""con""" , _a )
_A : int = self.to_sql_kwargs.pop("""index""" , _a )
_A : List[str] = self._write(index=_a , **self.to_sql_kwargs )
return written
def a__ ( self , _a ) -> Optional[int]:
_A , _A , _A : List[str] = args
_A : int = {**to_sql_kwargs, """if_exists""": """append"""} if offset > 0 else to_sql_kwargs
_A : str = query_table(
table=self.dataset.data , key=slice(_a , offset + self.batch_size ) , indices=self.dataset._indices , )
_A : Tuple = batch.to_pandas()
_A : Union[str, Any] = df.to_sql(self.name , self.con , index=_a , **_a )
return num_rows or len(_a )
def a__ ( self , _a , **_a ) -> int:
_A : Any = 0
if self.num_proc is None or self.num_proc == 1:
for offset in logging.tqdm(
range(0 , len(self.dataset ) , self.batch_size ) , unit="""ba""" , disable=not logging.is_progress_bar_enabled() , desc="""Creating SQL from Arrow format""" , ):
written += self._batch_sql((offset, index, to_sql_kwargs) )
else:
_A , _A : Tuple = len(self.dataset ), self.batch_size
with multiprocessing.Pool(self.num_proc ) as pool:
for num_rows in logging.tqdm(
pool.imap(
self._batch_sql , [(offset, index, to_sql_kwargs) for offset in range(0 , _a , _a )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit="""ba""" , disable=not logging.is_progress_bar_enabled() , desc="""Creating SQL from Arrow format""" , ):
written += num_rows
return written
| 26 | 1 |
from ...configuration_utils import PretrainedConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"transfo-xl-wt103": "https://huggingface.co/transfo-xl-wt103/resolve/main/config.json",
}
class lowercase ( UpperCamelCase__ ):
_a = "transfo-xl"
_a = ["mems"]
_a = {
"n_token": "vocab_size",
"hidden_size": "d_model",
"num_attention_heads": "n_head",
"num_hidden_layers": "n_layer",
}
def __init__( self , _a=26_7735 , _a=[2_0000, 4_0000, 20_0000] , _a=1024 , _a=1024 , _a=16 , _a=64 , _a=4096 , _a=4 , _a=False , _a=18 , _a=1600 , _a=1000 , _a=True , _a=True , _a=0 , _a=-1 , _a=True , _a=0.1 , _a=0.0 , _a=True , _a="normal" , _a=0.01 , _a=0.01 , _a=0.02 , _a=1e-5 , _a=0 , **_a , ) -> Tuple:
_A : Union[str, Any] = vocab_size
_A : Union[str, Any] = []
self.cutoffs.extend(_a )
if proj_share_all_but_first:
_A : Union[str, Any] = [False] + [True] * len(self.cutoffs )
else:
_A : Dict = [False] + [False] * len(self.cutoffs )
_A : List[str] = d_model
_A : Dict = d_embed
_A : Dict = d_head
_A : Optional[int] = d_inner
_A : Tuple = div_val
_A : Union[str, Any] = pre_lnorm
_A : Union[str, Any] = n_layer
_A : str = n_head
_A : Optional[Any] = mem_len
_A : Tuple = same_length
_A : Optional[int] = attn_type
_A : Any = clamp_len
_A : Dict = sample_softmax
_A : Any = adaptive
_A : List[str] = dropout
_A : List[str] = dropatt
_A : Union[str, Any] = untie_r
_A : Optional[int] = init
_A : Union[str, Any] = init_range
_A : Tuple = proj_init_std
_A : int = init_std
_A : List[str] = layer_norm_epsilon
super().__init__(eos_token_id=_a , **_a )
@property
def a__ ( self ) -> int:
# Message copied from Transformer-XL documentation
logger.info(F'''The model {self.model_type} is one of the few models that has no sequence length limit.''' )
return -1
@max_position_embeddings.setter
def a__ ( self , _a ) -> Tuple:
# Message copied from Transformer-XL documentation
raise NotImplementedError(
F'''The model {self.model_type} is one of the few models that has no sequence length limit.''' )
| 26 |
from ...configuration_utils import PretrainedConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/config.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/config.json"
# See all FNet models at https://huggingface.co/models?filter=fnet
}
class lowercase ( UpperCamelCase__ ):
_a = "fnet"
def __init__( self , _a=3_2000 , _a=768 , _a=12 , _a=3072 , _a="gelu_new" , _a=0.1 , _a=512 , _a=4 , _a=0.02 , _a=1e-12 , _a=False , _a=512 , _a=3 , _a=1 , _a=2 , **_a , ) -> int:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Any = vocab_size
_A : str = max_position_embeddings
_A : Optional[Any] = hidden_size
_A : List[str] = num_hidden_layers
_A : List[str] = intermediate_size
_A : List[Any] = hidden_act
_A : List[str] = hidden_dropout_prob
_A : List[str] = initializer_range
_A : List[Any] = type_vocab_size
_A : List[Any] = layer_norm_eps
_A : List[str] = use_tpu_fourier_optimizations
_A : str = tpu_short_seq_length
| 26 | 1 |
import torch
from torch import nn
from ...configuration_utils import ConfigMixin, register_to_config
from ...models import ModelMixin
class lowercase ( UpperCamelCase__,UpperCamelCase__ ):
@register_to_config
def __init__( self , *,
_a = 4 , _a = 768 , _a , _a , ) -> str:
super().__init__()
_A : List[Any] = nn.Parameter(torch.zeros(_a ) )
# parameters for additional clip time embeddings
_A : Union[str, Any] = nn.Linear(_a , _a )
_A : List[str] = nn.Linear(_a , _a )
# parameters for encoder hidden states
_A : List[str] = clip_extra_context_tokens
_A : Optional[int] = nn.Linear(
_a , self.clip_extra_context_tokens * cross_attention_dim )
_A : Union[str, Any] = nn.Linear(_a , _a )
_A : Tuple = nn.LayerNorm(_a )
def a__ ( self , *, _a , _a , _a , _a ) -> Union[str, Any]:
if do_classifier_free_guidance:
# Add the classifier free guidance embeddings to the image embeddings
_A : Union[str, Any] = image_embeddings.shape[0]
_A : Optional[Any] = self.learned_classifier_free_guidance_embeddings.unsqueeze(0 )
_A : str = classifier_free_guidance_embeddings.expand(
_a , -1 )
_A : List[str] = torch.cat([classifier_free_guidance_embeddings, image_embeddings] , dim=0 )
# The image embeddings batch size and the text embeddings batch size are equal
assert image_embeddings.shape[0] == prompt_embeds.shape[0]
_A : Any = prompt_embeds.shape[0]
# "Specifically, we modify the architecture described in Nichol et al. (2021) by projecting and
# adding CLIP embeddings to the existing timestep embedding, ...
_A : Any = self.embedding_proj(_a )
_A : Dict = self.clip_image_embeddings_project_to_time_embeddings(_a )
_A : List[str] = time_projected_image_embeddings + time_projected_prompt_embeds
# ... and by projecting CLIP embeddings into four
# extra tokens of context that are concatenated to the sequence of outputs from the GLIDE text encoder"
_A : Optional[int] = self.clip_extra_context_tokens_proj(_a )
_A : Union[str, Any] = clip_extra_context_tokens.reshape(_a , -1 , self.clip_extra_context_tokens )
_A : str = clip_extra_context_tokens.permute(0 , 2 , 1 )
_A : Dict = self.encoder_hidden_states_proj(_a )
_A : Any = self.text_encoder_hidden_states_norm(_a )
_A : Union[str, Any] = torch.cat([clip_extra_context_tokens, text_encoder_hidden_states] , dim=1 )
return text_encoder_hidden_states, additive_clip_time_embeddings
| 26 |
def lowerCAmelCase_ ( snake_case_ ):
if n_term == "":
return []
_A : list = []
for temp in range(int(snake_case_ ) ):
series.append(f'''1/{temp + 1}''' if series else """1""" )
return series
if __name__ == "__main__":
_snake_case = input("Enter the last number (nth term) of the Harmonic Series")
print("Formula of Harmonic Series => 1+1/2+1/3 ..... 1/n")
print(harmonic_series(nth_term))
| 26 | 1 |
from math import pi, sqrt
def lowerCAmelCase_ ( snake_case_ ):
if num <= 0:
raise ValueError("""math domain error""" )
if num > 1_71.5:
raise OverflowError("""math range error""" )
elif num - int(snake_case_ ) not in (0, 0.5):
raise NotImplementedError("""num must be an integer or a half-integer""" )
elif num == 0.5:
return sqrt(snake_case_ )
else:
return 1.0 if num == 1 else (num - 1) * gamma(num - 1 )
def lowerCAmelCase_ ( ):
assert gamma(0.5 ) == sqrt(snake_case_ )
assert gamma(1 ) == 1.0
assert gamma(2 ) == 1.0
if __name__ == "__main__":
from doctest import testmod
testmod()
_snake_case = 1.0
while num:
_snake_case = float(input("Gamma of: "))
print(f"""gamma({num}) = {gamma(num)}""")
print("\nEnter 0 to exit...")
| 26 |
import importlib
import json
import os
from collections import OrderedDict
from typing import Dict, Optional, Union
# Build the list of all feature extractors
from ...configuration_utils import PretrainedConfig
from ...dynamic_module_utils import get_class_from_dynamic_module, resolve_trust_remote_code
from ...feature_extraction_utils import FeatureExtractionMixin
from ...utils import CONFIG_NAME, FEATURE_EXTRACTOR_NAME, get_file_from_repo, logging
from .auto_factory import _LazyAutoMapping
from .configuration_auto import (
CONFIG_MAPPING_NAMES,
AutoConfig,
model_type_to_module_name,
replace_list_option_in_docstrings,
)
_snake_case = logging.get_logger(__name__)
_snake_case = OrderedDict(
[
("audio-spectrogram-transformer", "ASTFeatureExtractor"),
("beit", "BeitFeatureExtractor"),
("chinese_clip", "ChineseCLIPFeatureExtractor"),
("clap", "ClapFeatureExtractor"),
("clip", "CLIPFeatureExtractor"),
("clipseg", "ViTFeatureExtractor"),
("conditional_detr", "ConditionalDetrFeatureExtractor"),
("convnext", "ConvNextFeatureExtractor"),
("cvt", "ConvNextFeatureExtractor"),
("data2vec-audio", "Wav2Vec2FeatureExtractor"),
("data2vec-vision", "BeitFeatureExtractor"),
("deformable_detr", "DeformableDetrFeatureExtractor"),
("deit", "DeiTFeatureExtractor"),
("detr", "DetrFeatureExtractor"),
("dinat", "ViTFeatureExtractor"),
("donut-swin", "DonutFeatureExtractor"),
("dpt", "DPTFeatureExtractor"),
("encodec", "EncodecFeatureExtractor"),
("flava", "FlavaFeatureExtractor"),
("glpn", "GLPNFeatureExtractor"),
("groupvit", "CLIPFeatureExtractor"),
("hubert", "Wav2Vec2FeatureExtractor"),
("imagegpt", "ImageGPTFeatureExtractor"),
("layoutlmv2", "LayoutLMv2FeatureExtractor"),
("layoutlmv3", "LayoutLMv3FeatureExtractor"),
("levit", "LevitFeatureExtractor"),
("maskformer", "MaskFormerFeatureExtractor"),
("mctct", "MCTCTFeatureExtractor"),
("mobilenet_v1", "MobileNetV1FeatureExtractor"),
("mobilenet_v2", "MobileNetV2FeatureExtractor"),
("mobilevit", "MobileViTFeatureExtractor"),
("nat", "ViTFeatureExtractor"),
("owlvit", "OwlViTFeatureExtractor"),
("perceiver", "PerceiverFeatureExtractor"),
("poolformer", "PoolFormerFeatureExtractor"),
("regnet", "ConvNextFeatureExtractor"),
("resnet", "ConvNextFeatureExtractor"),
("segformer", "SegformerFeatureExtractor"),
("sew", "Wav2Vec2FeatureExtractor"),
("sew-d", "Wav2Vec2FeatureExtractor"),
("speech_to_text", "Speech2TextFeatureExtractor"),
("speecht5", "SpeechT5FeatureExtractor"),
("swiftformer", "ViTFeatureExtractor"),
("swin", "ViTFeatureExtractor"),
("swinv2", "ViTFeatureExtractor"),
("table-transformer", "DetrFeatureExtractor"),
("timesformer", "VideoMAEFeatureExtractor"),
("tvlt", "TvltFeatureExtractor"),
("unispeech", "Wav2Vec2FeatureExtractor"),
("unispeech-sat", "Wav2Vec2FeatureExtractor"),
("van", "ConvNextFeatureExtractor"),
("videomae", "VideoMAEFeatureExtractor"),
("vilt", "ViltFeatureExtractor"),
("vit", "ViTFeatureExtractor"),
("vit_mae", "ViTFeatureExtractor"),
("vit_msn", "ViTFeatureExtractor"),
("wav2vec2", "Wav2Vec2FeatureExtractor"),
("wav2vec2-conformer", "Wav2Vec2FeatureExtractor"),
("wavlm", "Wav2Vec2FeatureExtractor"),
("whisper", "WhisperFeatureExtractor"),
("xclip", "CLIPFeatureExtractor"),
("yolos", "YolosFeatureExtractor"),
]
)
_snake_case = _LazyAutoMapping(CONFIG_MAPPING_NAMES, FEATURE_EXTRACTOR_MAPPING_NAMES)
def lowerCAmelCase_ ( snake_case_ ):
for module_name, extractors in FEATURE_EXTRACTOR_MAPPING_NAMES.items():
if class_name in extractors:
_A : List[str] = model_type_to_module_name(snake_case_ )
_A : List[Any] = importlib.import_module(f'''.{module_name}''',"""transformers.models""" )
try:
return getattr(snake_case_,snake_case_ )
except AttributeError:
continue
for _, extractor in FEATURE_EXTRACTOR_MAPPING._extra_content.items():
if getattr(snake_case_,"""__name__""",snake_case_ ) == class_name:
return extractor
# We did not fine the class, but maybe it's because a dep is missing. In that case, the class will be in the main
# init and we return the proper dummy to get an appropriate error message.
_A : List[Any] = importlib.import_module("""transformers""" )
if hasattr(snake_case_,snake_case_ ):
return getattr(snake_case_,snake_case_ )
return None
def lowerCAmelCase_ ( snake_case_,snake_case_ = None,snake_case_ = False,snake_case_ = False,snake_case_ = None,snake_case_ = None,snake_case_ = None,snake_case_ = False,**snake_case_,):
_A : Optional[int] = get_file_from_repo(
snake_case_,snake_case_,cache_dir=snake_case_,force_download=snake_case_,resume_download=snake_case_,proxies=snake_case_,use_auth_token=snake_case_,revision=snake_case_,local_files_only=snake_case_,)
if resolved_config_file is None:
logger.info(
"""Could not locate the feature extractor configuration file, will try to use the model config instead.""" )
return {}
with open(snake_case_,encoding="""utf-8""" ) as reader:
return json.load(snake_case_ )
class lowercase :
def __init__( self ) -> List[Any]:
raise EnvironmentError(
"""AutoFeatureExtractor is designed to be instantiated """
"""using the `AutoFeatureExtractor.from_pretrained(pretrained_model_name_or_path)` method.""" )
@classmethod
@replace_list_option_in_docstrings(_a )
def a__ ( cls , _a , **_a ) -> Any:
_A : Tuple = kwargs.pop("""config""" , _a )
_A : Tuple = kwargs.pop("""trust_remote_code""" , _a )
_A : List[Any] = True
_A , _A : Tuple = FeatureExtractionMixin.get_feature_extractor_dict(_a , **_a )
_A : Tuple = config_dict.get("""feature_extractor_type""" , _a )
_A : int = None
if "AutoFeatureExtractor" in config_dict.get("""auto_map""" , {} ):
_A : Optional[int] = config_dict["""auto_map"""]["""AutoFeatureExtractor"""]
# If we don't find the feature extractor class in the feature extractor config, let's try the model config.
if feature_extractor_class is None and feature_extractor_auto_map is None:
if not isinstance(_a , _a ):
_A : int = AutoConfig.from_pretrained(_a , **_a )
# It could be in `config.feature_extractor_type``
_A : Optional[int] = getattr(_a , """feature_extractor_type""" , _a )
if hasattr(_a , """auto_map""" ) and "AutoFeatureExtractor" in config.auto_map:
_A : Tuple = config.auto_map["""AutoFeatureExtractor"""]
if feature_extractor_class is not None:
_A : Optional[Any] = feature_extractor_class_from_name(_a )
_A : List[Any] = feature_extractor_auto_map is not None
_A : Union[str, Any] = feature_extractor_class is not None or type(_a ) in FEATURE_EXTRACTOR_MAPPING
_A : Optional[int] = resolve_trust_remote_code(
_a , _a , _a , _a )
if has_remote_code and trust_remote_code:
_A : Dict = get_class_from_dynamic_module(
_a , _a , **_a )
_A : str = kwargs.pop("""code_revision""" , _a )
if os.path.isdir(_a ):
feature_extractor_class.register_for_auto_class()
return feature_extractor_class.from_dict(_a , **_a )
elif feature_extractor_class is not None:
return feature_extractor_class.from_dict(_a , **_a )
# Last try: we use the FEATURE_EXTRACTOR_MAPPING.
elif type(_a ) in FEATURE_EXTRACTOR_MAPPING:
_A : Dict = FEATURE_EXTRACTOR_MAPPING[type(_a )]
return feature_extractor_class.from_dict(_a , **_a )
raise ValueError(
F'''Unrecognized feature extractor in {pretrained_model_name_or_path}. Should have a '''
F'''`feature_extractor_type` key in its {FEATURE_EXTRACTOR_NAME} of {CONFIG_NAME}, or one of the following '''
F'''`model_type` keys in its {CONFIG_NAME}: {", ".join(c for c in FEATURE_EXTRACTOR_MAPPING_NAMES.keys() )}''' )
@staticmethod
def a__ ( _a , _a ) -> Optional[int]:
FEATURE_EXTRACTOR_MAPPING.register(_a , _a )
| 26 | 1 |
import unittest
from transformers import load_tool
from .test_tools_common import ToolTesterMixin
class lowercase ( unittest.TestCase,UpperCamelCase__ ):
def a__ ( self ) -> Union[str, Any]:
_A : Union[str, Any] = load_tool("""text-classification""" )
self.tool.setup()
_A : Optional[Any] = load_tool("""text-classification""" , remote=_a )
def a__ ( self ) -> str:
_A : str = self.tool("""That's quite cool""" , ["""positive""", """negative"""] )
self.assertEqual(_a , """positive""" )
def a__ ( self ) -> Union[str, Any]:
_A : Tuple = self.remote_tool("""That's quite cool""" , ["""positive""", """negative"""] )
self.assertEqual(_a , """positive""" )
def a__ ( self ) -> Optional[Any]:
_A : Union[str, Any] = self.tool(text="""That's quite cool""" , labels=["""positive""", """negative"""] )
self.assertEqual(_a , """positive""" )
def a__ ( self ) -> Dict:
_A : List[Any] = self.remote_tool(text="""That's quite cool""" , labels=["""positive""", """negative"""] )
self.assertEqual(_a , """positive""" )
| 26 |
import unittest
import numpy as np
from transformers.testing_utils import is_flaky, require_torch, require_vision
from transformers.utils import is_torch_available, is_vision_available
from ...test_image_processing_common import ImageProcessingSavingTestMixin, prepare_image_inputs
if is_torch_available():
import torch
if is_vision_available():
from PIL import Image
from transformers import DonutImageProcessor
class lowercase ( unittest.TestCase ):
def __init__( self , _a , _a=7 , _a=3 , _a=18 , _a=30 , _a=400 , _a=True , _a=None , _a=True , _a=False , _a=True , _a=True , _a=[0.5, 0.5, 0.5] , _a=[0.5, 0.5, 0.5] , ) -> Dict:
_A : str = parent
_A : int = batch_size
_A : Optional[int] = num_channels
_A : List[Any] = image_size
_A : int = min_resolution
_A : Optional[int] = max_resolution
_A : Any = do_resize
_A : List[str] = size if size is not None else {"""height""": 18, """width""": 20}
_A : Optional[int] = do_thumbnail
_A : str = do_align_axis
_A : List[Any] = do_pad
_A : Optional[Any] = do_normalize
_A : Tuple = image_mean
_A : List[str] = image_std
def a__ ( self ) -> Optional[int]:
return {
"do_resize": self.do_resize,
"size": self.size,
"do_thumbnail": self.do_thumbnail,
"do_align_long_axis": self.do_align_axis,
"do_pad": self.do_pad,
"do_normalize": self.do_normalize,
"image_mean": self.image_mean,
"image_std": self.image_std,
}
@require_torch
@require_vision
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = DonutImageProcessor if is_vision_available() else None
def a__ ( self ) -> Optional[int]:
_A : List[str] = DonutImageProcessingTester(self )
@property
def a__ ( self ) -> List[Any]:
return self.image_processor_tester.prepare_image_processor_dict()
def a__ ( self ) -> Optional[Any]:
_A : Union[str, Any] = self.image_processing_class(**self.image_processor_dict )
self.assertTrue(hasattr(_a , """do_resize""" ) )
self.assertTrue(hasattr(_a , """size""" ) )
self.assertTrue(hasattr(_a , """do_thumbnail""" ) )
self.assertTrue(hasattr(_a , """do_align_long_axis""" ) )
self.assertTrue(hasattr(_a , """do_pad""" ) )
self.assertTrue(hasattr(_a , """do_normalize""" ) )
self.assertTrue(hasattr(_a , """image_mean""" ) )
self.assertTrue(hasattr(_a , """image_std""" ) )
def a__ ( self ) -> List[Any]:
_A : Union[str, Any] = self.image_processing_class.from_dict(self.image_processor_dict )
self.assertEqual(image_processor.size , {"""height""": 18, """width""": 20} )
_A : Optional[int] = self.image_processing_class.from_dict(self.image_processor_dict , size=42 )
self.assertEqual(image_processor.size , {"""height""": 42, """width""": 42} )
# Previous config had dimensions in (width, height) order
_A : List[str] = self.image_processing_class.from_dict(self.image_processor_dict , size=(42, 84) )
self.assertEqual(image_processor.size , {"""height""": 84, """width""": 42} )
def a__ ( self ) -> Union[str, Any]:
pass
@is_flaky()
def a__ ( self ) -> Optional[int]:
# Initialize image_processing
_A : List[Any] = self.image_processing_class(**self.image_processor_dict )
# create random PIL images
_A : Dict = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a )
for image in image_inputs:
self.assertIsInstance(_a , Image.Image )
# Test not batched input
_A : Union[str, Any] = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : Any = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
@is_flaky()
def a__ ( self ) -> Dict:
# Initialize image_processing
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random numpy tensors
_A : Tuple = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , numpify=_a )
for image in image_inputs:
self.assertIsInstance(_a , np.ndarray )
# Test not batched input
_A : int = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : List[str] = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
@is_flaky()
def a__ ( self ) -> Optional[int]:
# Initialize image_processing
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random PyTorch tensors
_A : Optional[int] = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , torchify=_a )
for image in image_inputs:
self.assertIsInstance(_a , torch.Tensor )
# Test not batched input
_A : Any = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : str = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
| 26 | 1 |
class lowercase :
def __init__( self ) -> List[str]:
_A : List[Any] = {}
def a__ ( self ) -> None:
print(self.vertex )
for i in self.vertex:
print(_a , """ -> """ , """ -> """.join([str(_a ) for j in self.vertex[i]] ) )
def a__ ( self , _a , _a ) -> None:
# check if vertex is already present,
if from_vertex in self.vertex:
self.vertex[from_vertex].append(_a )
else:
# else make a new vertex
_A : Dict = [to_vertex]
def a__ ( self ) -> None:
# visited array for storing already visited nodes
_A : List[Any] = [False] * len(self.vertex )
# call the recursive helper function
for i in range(len(self.vertex ) ):
if not visited[i]:
self.dfs_recursive(_a , _a )
def a__ ( self , _a , _a ) -> None:
# mark start vertex as visited
_A : Tuple = True
print(_a , end=""" """ )
# Recur for all the vertices that are adjacent to this node
for i in self.vertex:
if not visited[i]:
self.dfs_recursive(_a , _a )
if __name__ == "__main__":
_snake_case = Graph()
g.add_edge(0, 1)
g.add_edge(0, 2)
g.add_edge(1, 2)
g.add_edge(2, 0)
g.add_edge(2, 3)
g.add_edge(3, 3)
g.print_graph()
print("DFS:")
g.dfs()
# OUTPUT:
# 0 -> 1 -> 2
# 1 -> 2
# 2 -> 0 -> 3
# 3 -> 3
# DFS:
# 0 1 2 3
| 26 |
from __future__ import annotations
import numpy as np
def lowerCAmelCase_ ( snake_case_ ):
return np.maximum(0,snake_case_ )
if __name__ == "__main__":
print(np.array(relu([-1, 0, 5]))) # --> [0, 0, 5]
| 26 | 1 |
import gc
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
PNDMScheduler,
StableDiffusionLDMaDPipeline,
UNetaDConditionModel,
)
from diffusers.utils import nightly, slow, torch_device
from diffusers.utils.testing_utils import enable_full_determinism, require_torch_gpu
from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
enable_full_determinism()
class lowercase ( unittest.TestCase ):
_a = StableDiffusionLDMaDPipeline
_a = TEXT_TO_IMAGE_PARAMS
_a = TEXT_TO_IMAGE_BATCH_PARAMS
_a = TEXT_TO_IMAGE_IMAGE_PARAMS
def a__ ( self ) -> List[str]:
torch.manual_seed(0 )
_A : Optional[Any] = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=2 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , cross_attention_dim=32 , )
_A : Any = DDIMScheduler(
beta_start=0.00085 , beta_end=0.012 , beta_schedule="""scaled_linear""" , clip_sample=_a , set_alpha_to_one=_a , )
torch.manual_seed(0 )
_A : Union[str, Any] = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=6 , out_channels=6 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , )
torch.manual_seed(0 )
_A : Tuple = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-05 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1000 , )
_A : Optional[int] = CLIPTextModel(_a )
_A : Any = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" )
_A : Optional[Any] = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""safety_checker""": None,
"""feature_extractor""": None,
}
return components
def a__ ( self , _a , _a=0 ) -> Tuple:
if str(_a ).startswith("""mps""" ):
_A : Tuple = torch.manual_seed(_a )
else:
_A : List[str] = torch.Generator(device=_a ).manual_seed(_a )
_A : str = {
"""prompt""": """A painting of a squirrel eating a burger""",
"""generator""": generator,
"""num_inference_steps""": 2,
"""guidance_scale""": 6.0,
"""output_type""": """numpy""",
}
return inputs
def a__ ( self ) -> List[Any]:
_A : Dict = """cpu""" # ensure determinism for the device-dependent torch.Generator
_A : Optional[int] = self.get_dummy_components()
_A : Union[str, Any] = StableDiffusionLDMaDPipeline(**_a )
_A : Optional[int] = ldmad_pipe.to(_a )
ldmad_pipe.set_progress_bar_config(disable=_a )
_A : Dict = self.get_dummy_inputs(_a )
_A : Optional[Any] = ldmad_pipe(**_a )
_A , _A : str = output.rgb, output.depth
_A : Union[str, Any] = rgb[0, -3:, -3:, -1]
_A : Tuple = depth[0, -3:, -1]
assert rgb.shape == (1, 64, 64, 3)
assert depth.shape == (1, 64, 64)
_A : Any = np.array(
[0.37338176, 0.70247, 0.74203193, 0.51643604, 0.58256793, 0.60932136, 0.4181095, 0.48355877, 0.46535262] )
_A : Dict = np.array([103.46727, 85.812004, 87.849236] )
assert np.abs(image_slice_rgb.flatten() - expected_slice_rgb ).max() < 1e-2
assert np.abs(image_slice_depth.flatten() - expected_slice_depth ).max() < 1e-2
def a__ ( self ) -> List[str]:
_A : int = self.get_dummy_components()
_A : Optional[Any] = StableDiffusionLDMaDPipeline(**_a )
_A : Any = ldmad_pipe.to(_a )
ldmad_pipe.set_progress_bar_config(disable=_a )
_A : str = self.get_dummy_inputs(_a )
_A : List[str] = 3 * [inputs["""prompt"""]]
# forward
_A : Tuple = ldmad_pipe(**_a )
_A , _A : Optional[int] = output.rgb, output.depth
_A : Dict = rgb_slice_a[0, -3:, -3:, -1]
_A : int = depth_slice_a[0, -3:, -1]
_A : Optional[Any] = self.get_dummy_inputs(_a )
_A : Optional[Any] = 3 * [inputs.pop("""prompt""" )]
_A : Optional[Any] = ldmad_pipe.tokenizer(
_a , padding="""max_length""" , max_length=ldmad_pipe.tokenizer.model_max_length , truncation=_a , return_tensors="""pt""" , )
_A : Optional[Any] = text_inputs["""input_ids"""].to(_a )
_A : Dict = ldmad_pipe.text_encoder(_a )[0]
_A : Dict = prompt_embeds
# forward
_A : Dict = ldmad_pipe(**_a )
_A , _A : Optional[Any] = output.rgb, output.depth
_A : Any = rgb_slice_a[0, -3:, -3:, -1]
_A : Any = depth_slice_a[0, -3:, -1]
assert np.abs(rgb_slice_a.flatten() - rgb_slice_a.flatten() ).max() < 1e-4
assert np.abs(depth_slice_a.flatten() - depth_slice_a.flatten() ).max() < 1e-4
def a__ ( self ) -> Tuple:
_A : Optional[int] = """cpu""" # ensure determinism for the device-dependent torch.Generator
_A : List[str] = self.get_dummy_components()
_A : Optional[int] = PNDMScheduler(skip_prk_steps=_a )
_A : Optional[int] = StableDiffusionLDMaDPipeline(**_a )
_A : int = ldmad_pipe.to(_a )
ldmad_pipe.set_progress_bar_config(disable=_a )
_A : Any = self.get_dummy_inputs(_a )
_A : Any = """french fries"""
_A : List[Any] = ldmad_pipe(**_a , negative_prompt=_a )
_A , _A : Dict = output.rgb, output.depth
_A : Dict = rgb[0, -3:, -3:, -1]
_A : int = depth[0, -3:, -1]
assert rgb.shape == (1, 64, 64, 3)
assert depth.shape == (1, 64, 64)
_A : int = np.array(
[0.37044, 0.71811503, 0.7223251, 0.48603675, 0.5638391, 0.6364948, 0.42833704, 0.4901315, 0.47926217] )
_A : str = np.array([107.84738, 84.62802, 89.962135] )
assert np.abs(rgb_slice.flatten() - expected_slice_rgb ).max() < 1e-2
assert np.abs(depth_slice.flatten() - expected_slice_depth ).max() < 1e-2
@slow
@require_torch_gpu
class lowercase ( unittest.TestCase ):
def a__ ( self ) -> List[str]:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self , _a , _a="cpu" , _a=torch.floataa , _a=0 ) -> Any:
_A : Tuple = torch.Generator(device=_a ).manual_seed(_a )
_A : Union[str, Any] = np.random.RandomState(_a ).standard_normal((1, 4, 64, 64) )
_A : Dict = torch.from_numpy(_a ).to(device=_a , dtype=_a )
_A : int = {
"""prompt""": """a photograph of an astronaut riding a horse""",
"""latents""": latents,
"""generator""": generator,
"""num_inference_steps""": 3,
"""guidance_scale""": 7.5,
"""output_type""": """numpy""",
}
return inputs
def a__ ( self ) -> str:
_A : List[str] = StableDiffusionLDMaDPipeline.from_pretrained("""Intel/ldm3d""" )
_A : List[str] = ldmad_pipe.to(_a )
ldmad_pipe.set_progress_bar_config(disable=_a )
_A : Optional[Any] = self.get_inputs(_a )
_A : Optional[Any] = ldmad_pipe(**_a )
_A , _A : Union[str, Any] = output.rgb, output.depth
_A : Optional[Any] = rgb[0, -3:, -3:, -1].flatten()
_A : Dict = rgb[0, -3:, -1].flatten()
assert rgb.shape == (1, 512, 512, 3)
assert depth.shape == (1, 512, 512)
_A : List[str] = np.array(
[0.53805465, 0.56707305, 0.5486515, 0.57012236, 0.5814511, 0.56253487, 0.54843014, 0.55092263, 0.6459706] )
_A : Any = np.array(
[0.9263781, 0.6678672, 0.5486515, 0.92202145, 0.67831135, 0.56253487, 0.9241694, 0.7551478, 0.6459706] )
assert np.abs(rgb_slice - expected_slice_rgb ).max() < 3e-3
assert np.abs(depth_slice - expected_slice_depth ).max() < 3e-3
@nightly
@require_torch_gpu
class lowercase ( unittest.TestCase ):
def a__ ( self ) -> Tuple:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self , _a , _a="cpu" , _a=torch.floataa , _a=0 ) -> Any:
_A : List[Any] = torch.Generator(device=_a ).manual_seed(_a )
_A : Optional[Any] = np.random.RandomState(_a ).standard_normal((1, 4, 64, 64) )
_A : int = torch.from_numpy(_a ).to(device=_a , dtype=_a )
_A : int = {
"""prompt""": """a photograph of an astronaut riding a horse""",
"""latents""": latents,
"""generator""": generator,
"""num_inference_steps""": 50,
"""guidance_scale""": 7.5,
"""output_type""": """numpy""",
}
return inputs
def a__ ( self ) -> Dict:
_A : List[str] = StableDiffusionLDMaDPipeline.from_pretrained("""Intel/ldm3d""" ).to(_a )
ldmad_pipe.set_progress_bar_config(disable=_a )
_A : List[Any] = self.get_inputs(_a )
_A : Dict = ldmad_pipe(**_a )
_A , _A : Dict = output.rgb, output.depth
_A : Any = 0.495586
_A : str = 0.33795515
_A : List[Any] = 112.48518
_A : Optional[int] = 98.489746
assert np.abs(expected_rgb_mean - rgb.mean() ) < 1e-3
assert np.abs(expected_rgb_std - rgb.std() ) < 1e-3
assert np.abs(expected_depth_mean - depth.mean() ) < 1e-3
assert np.abs(expected_depth_std - depth.std() ) < 1e-3
def a__ ( self ) -> Optional[int]:
_A : Dict = StableDiffusionLDMaDPipeline.from_pretrained("""Intel/ldm3d-4c""" ).to(_a )
ldmad_pipe.set_progress_bar_config(disable=_a )
_A : Any = self.get_inputs(_a )
_A : int = ldmad_pipe(**_a )
_A , _A : str = output.rgb, output.depth
_A : Any = 0.4194127
_A : int = 0.35375586
_A : int = 0.5638502
_A : Tuple = 0.34686103
assert rgb.shape == (1, 512, 512, 3)
assert depth.shape == (1, 512, 512, 1)
assert np.abs(expected_rgb_mean - rgb.mean() ) < 1e-3
assert np.abs(expected_rgb_std - rgb.std() ) < 1e-3
assert np.abs(expected_depth_mean - depth.mean() ) < 1e-3
assert np.abs(expected_depth_std - depth.std() ) < 1e-3
| 26 |
import argparse
import shutil
import time
from json import JSONDecodeError
from logging import getLogger
from pathlib import Path
from typing import Dict, List
import torch
from torch.utils.data import DataLoader
from tqdm import tqdm
from transformers import AutoModelForSeqaSeqLM, AutoTokenizer
from utils import (
SeqaSeqDataset,
calculate_bleu,
calculate_rouge,
chunks,
lmap,
load_json,
parse_numeric_n_bool_cl_kwargs,
save_json,
use_task_specific_params,
write_txt_file,
)
_snake_case = getLogger(__name__)
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ = 8,snake_case_ = 1024,snake_case_="val",snake_case_=None,snake_case_=False,snake_case_="summarization",snake_case_=None,snake_case_=1,snake_case_ = None,snake_case_="",**snake_case_,):
_A : Dict = str(snake_case_ )
assert local_rank is not None
torch.distributed.init_process_group(backend="""nccl""",rank=snake_case_ )
_A : Tuple = Path(snake_case_ )
_A : List[Any] = save_dir.joinpath(f'''rank_{local_rank}_output.json''' )
torch.cuda.set_device(snake_case_ )
_A : Union[str, Any] = AutoModelForSeqaSeqLM.from_pretrained(snake_case_ ).cuda()
if fpaa:
_A : Any = model.half()
# determine if we need to increase num_beams
use_task_specific_params(snake_case_,snake_case_ ) # update config with task specific params
_A : str = generate_kwargs.pop("""num_beams""",model.config.num_beams ) # AttributeError risk?
if num_return_sequences > num_beams:
_A : int = num_return_sequences
_A : Optional[Any] = AutoTokenizer.from_pretrained(snake_case_ )
logger.info(f'''Inferred tokenizer type: {tokenizer.__class__}''' ) # if this is wrong, check config.model_type.
if max_source_length is None:
_A : Optional[int] = tokenizer.model_max_length
if prefix is None:
_A : Tuple = prefix or getattr(model.config,"""prefix""","""""" ) or """"""
_A : Optional[int] = SeqaSeqDataset(
snake_case_,snake_case_,snake_case_,max_target_length=1024,type_path=snake_case_,n_obs=snake_case_,prefix=snake_case_,**snake_case_,)
# I set shuffle=True for a more accurate progress bar.
# If all the longest samples are first, the prog bar estimate is too high at the beginning.
_A : Optional[int] = ds.make_sortish_sampler(snake_case_,distributed=snake_case_,add_extra_examples=snake_case_,shuffle=snake_case_ )
_A : Dict = DataLoader(snake_case_,sampler=snake_case_,batch_size=snake_case_,collate_fn=ds.collate_fn )
_A : Optional[Any] = []
for batch in tqdm(snake_case_ ):
_A : Tuple = model.generate(
input_ids=batch["""input_ids"""].to(model.device ),attention_mask=batch["""attention_mask"""].to(model.device ),num_return_sequences=snake_case_,num_beams=snake_case_,**snake_case_,)
_A : Any = tokenizer.batch_decode(snake_case_,skip_special_tokens=snake_case_,clean_up_tokenization_spaces=snake_case_ )
_A : Dict = batch["""ids"""]
if num_return_sequences > 1:
_A : Any = chunks(snake_case_,snake_case_ ) # batch size chunks, each of size num_return_seq
for i, pred in enumerate(snake_case_ ):
results.append({"""pred""": pred, """id""": ids[i].item()} )
save_json(snake_case_,snake_case_ )
return results, sampler.num_replicas
def lowerCAmelCase_ ( ):
_A : Tuple = argparse.ArgumentParser(
epilog="""Unspecified args like --num_beams=2 --decoder_start_token_id=4 are passed to model.generate""" )
parser.add_argument("""--data_dir""",type=snake_case_,help="""like cnn_dm/test.source""" )
parser.add_argument(
"""--model_name""",type=snake_case_,help="""like facebook/bart-large-cnn,t5-base, etc.""",default="""sshleifer/distilbart-xsum-12-3""",)
parser.add_argument("""--save_dir""",type=snake_case_,help="""where to save""",default="""tmp_gen""" )
parser.add_argument("""--max_source_length""",type=snake_case_,default=snake_case_ )
parser.add_argument(
"""--type_path""",type=snake_case_,default="""test""",help="""which subset to evaluate typically train/val/test""" )
parser.add_argument("""--task""",type=snake_case_,default="""summarization""",help="""used for task_specific_params + metrics""" )
parser.add_argument("""--bs""",type=snake_case_,default=8,required=snake_case_,help="""batch size""" )
parser.add_argument(
"""--local_rank""",type=snake_case_,default=-1,required=snake_case_,help="""should be passed by distributed.launch""" )
parser.add_argument(
"""--n_obs""",type=snake_case_,default=snake_case_,required=snake_case_,help="""How many observations. Defaults to all.""" )
parser.add_argument(
"""--num_return_sequences""",type=snake_case_,default=1,required=snake_case_,help="""How many sequences to return""" )
parser.add_argument(
"""--sync_timeout""",type=snake_case_,default=600,required=snake_case_,help="""How long should master process wait for other processes to finish.""",)
parser.add_argument("""--src_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument("""--tgt_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument(
"""--prefix""",type=snake_case_,required=snake_case_,default=snake_case_,help="""will be added to the begininng of src examples""" )
parser.add_argument("""--fp16""",action="""store_true""" )
parser.add_argument("""--debug""",action="""store_true""" )
_A : Union[str, Any] = time.time()
_A , _A : List[str] = parser.parse_known_args()
_A : List[str] = parse_numeric_n_bool_cl_kwargs(snake_case_ )
if generate_kwargs and args.local_rank <= 0:
print(f'''parsed the following generate kwargs: {generate_kwargs}''' )
_A : Dict = Path(args.save_dir + """_tmp""" )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ ) # this handles locking.
_A : int = list(json_save_dir.glob("""rank_*.json""" ) )
if intermediate_files:
raise ValueError(f'''Found files at {json_save_dir} please move or remove them.''' )
# In theory, a node could finish and save before another node hits this. If this happens, we can address later.
_A : Any = {}
if args.src_lang is not None:
_A : int = args.src_lang
if args.tgt_lang is not None:
_A : Dict = args.tgt_lang
Path(args.save_dir ).mkdir(exist_ok=snake_case_ )
_A , _A : str = eval_data_dir(
args.data_dir,snake_case_,args.model_name,type_path=args.type_path,bs=args.bs,fpaa=args.fpaa,task=args.task,local_rank=args.local_rank,n_obs=args.n_obs,max_source_length=args.max_source_length,num_return_sequences=args.num_return_sequences,prefix=args.prefix,dataset_kwargs=snake_case_,**snake_case_,)
if args.local_rank <= 0:
_A : List[Any] = Path(args.save_dir )
save_dir.mkdir(exist_ok=snake_case_ )
_A : Tuple = gather_results_from_each_node(snake_case_,snake_case_,args.sync_timeout )
_A : Optional[int] = combine_partial_results(snake_case_ )
if args.num_return_sequences > 1:
_A : Optional[Any] = save_dir.joinpath("""pseudolabel_results.json""" )
print(f'''Saving aggregated results at {save_path}, intermediate in {json_save_dir}/''' )
save_json(snake_case_,snake_case_ )
return
_A : List[str] = Path(args.data_dir ).joinpath(args.type_path + """.target""" )
with open(snake_case_ ) as f:
_A : int = [x.rstrip() for x in f.readlines()][: len(snake_case_ )]
# Calculate metrics, save metrics, and save _generations.txt
_A : Dict = """translation""" in args.task
_A : Optional[Any] = calculate_bleu if calc_bleu else calculate_rouge
_A : Tuple = """bleu""" if calc_bleu else """rouge"""
_A : Dict = score_fn(snake_case_,snake_case_ )
_A : List[Any] = len(snake_case_ )
_A : Optional[int] = time.time() - start_time
_A : Dict = round(runtime / metrics["""n_obs"""],4 )
_A : Dict = num_replicas
# TODO(@stas00): add whatever metadata to metrics
_A : Any = save_dir.joinpath(f'''{args.type_path}_{metric_name}.json''' )
save_json(snake_case_,snake_case_,indent=snake_case_ )
print(snake_case_ )
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}_generations.txt''' ) )
if args.debug:
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}.target''' ) )
else:
shutil.rmtree(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
_A : Dict = []
for partial_result in partial_results:
records.extend(snake_case_ )
_A : Optional[Any] = sorted(snake_case_,key=lambda snake_case_ : x["id"] )
_A : List[str] = [x["""pred"""] for x in records]
return preds
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
# WAIT FOR lots of .json files
_A : Optional[Any] = time.time()
logger.info("""waiting for all nodes to finish""" )
_A : List[str] = None
while (time.time() - start_wait) < timeout:
_A : str = list(save_dir.glob("""rank_*.json""" ) )
if len(snake_case_ ) < num_replicas:
continue
try:
# make sure all json files are fully saved
_A : List[str] = lmap(snake_case_,snake_case_ )
return json_data
except JSONDecodeError:
continue
else:
raise TimeoutError("""Rank 0 gave up on waiting for other processes""" )
# Unreachable
if __name__ == "__main__":
# Usage for MT:
run_generate()
| 26 | 1 |
from __future__ import annotations
from collections.abc import Iterator
from typing import Generic, TypeVar
_snake_case = TypeVar("T")
class lowercase ( Generic[T] ):
def __init__( self , _a ) -> str:
_A : Union[str, Any] = data
_A : Node[T] | None = None
def __str__( self ) -> str:
return F'''{self.data}'''
class lowercase ( Generic[T] ):
def __init__( self ) -> None:
_A : Node[T] | None = None
def __iter__( self ) -> Iterator[T]:
_A : Union[str, Any] = self.top
while node:
yield node.data
_A : Any = node.next
def __str__( self ) -> str:
return "->".join([str(_a ) for item in self] )
def __len__( self ) -> int:
return len(tuple(iter(self ) ) )
def a__ ( self ) -> bool:
return self.top is None
def a__ ( self , _a ) -> None:
_A : Any = Node(_a )
if not self.is_empty():
_A : int = self.top
_A : List[str] = node
def a__ ( self ) -> T:
if self.is_empty():
raise IndexError("""pop from empty stack""" )
assert isinstance(self.top , _a )
_A : int = self.top
_A : Any = self.top.next
return pop_node.data
def a__ ( self ) -> T:
if self.is_empty():
raise IndexError("""peek from empty stack""" )
assert self.top is not None
return self.top.data
def a__ ( self ) -> None:
_A : Dict = None
if __name__ == "__main__":
from doctest import testmod
testmod()
| 26 |
from __future__ import annotations
import unittest
from transformers import is_tf_available
from transformers.testing_utils import require_sentencepiece, require_tf, require_tokenizers, slow
if is_tf_available():
import numpy as np
import tensorflow as tf
from transformers import TFCamembertModel
@require_tf
@require_sentencepiece
@require_tokenizers
class lowercase ( unittest.TestCase ):
@slow
def a__ ( self ) -> Any:
_A : Tuple = TFCamembertModel.from_pretrained("""jplu/tf-camembert-base""" )
_A : List[Any] = tf.convert_to_tensor(
[[5, 121, 11, 660, 16, 730, 2_5543, 110, 83, 6]] , dtype=tf.intaa , ) # J'aime le camembert !"
_A : List[str] = model(_a )["""last_hidden_state"""]
_A : Union[str, Any] = tf.TensorShape((1, 10, 768) )
self.assertEqual(output.shape , _a )
# compare the actual values for a slice.
_A : List[Any] = tf.convert_to_tensor(
[[[-0.0254, 0.0235, 0.1027], [0.0606, -0.1811, -0.0418], [-0.1561, -0.1127, 0.2687]]] , dtype=tf.floataa , )
# camembert = torch.hub.load('pytorch/fairseq', 'camembert.v0')
# camembert.eval()
# expected_slice = roberta.model.forward(input_ids)[0][:, :3, :3].detach()
self.assertTrue(np.allclose(output[:, :3, :3].numpy() , expected_slice.numpy() , atol=1e-4 ) )
| 26 | 1 |
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_tokenizers_available, is_torch_available
_snake_case = {
"configuration_roc_bert": ["ROC_BERT_PRETRAINED_CONFIG_ARCHIVE_MAP", "RoCBertConfig"],
"tokenization_roc_bert": ["RoCBertTokenizer"],
}
try:
if not is_tokenizers_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
pass
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"ROC_BERT_PRETRAINED_MODEL_ARCHIVE_LIST",
"RoCBertForCausalLM",
"RoCBertForMaskedLM",
"RoCBertForMultipleChoice",
"RoCBertForPreTraining",
"RoCBertForQuestionAnswering",
"RoCBertForSequenceClassification",
"RoCBertForTokenClassification",
"RoCBertLayer",
"RoCBertModel",
"RoCBertPreTrainedModel",
"load_tf_weights_in_roc_bert",
]
if TYPE_CHECKING:
from .configuration_roc_bert import ROC_BERT_PRETRAINED_CONFIG_ARCHIVE_MAP, RoCBertConfig
from .tokenization_roc_bert import RoCBertTokenizer
try:
if not is_tokenizers_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
raise OptionalDependencyNotAvailable()
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_roc_bert import (
ROC_BERT_PRETRAINED_MODEL_ARCHIVE_LIST,
RoCBertForCausalLM,
RoCBertForMaskedLM,
RoCBertForMultipleChoice,
RoCBertForPreTraining,
RoCBertForQuestionAnswering,
RoCBertForSequenceClassification,
RoCBertForTokenClassification,
RoCBertLayer,
RoCBertModel,
RoCBertPreTrainedModel,
load_tf_weights_in_roc_bert,
)
else:
import sys
_snake_case = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
| 26 |
from dataclasses import dataclass
from typing import List, Optional, Union
import numpy as np
import PIL
import torch
from transformers import CLIPImageProcessor, CLIPVisionModel
from ...models import PriorTransformer
from ...pipelines import DiffusionPipeline
from ...schedulers import HeunDiscreteScheduler
from ...utils import (
BaseOutput,
is_accelerate_available,
logging,
randn_tensor,
replace_example_docstring,
)
from .renderer import ShapERenderer
_snake_case = logging.get_logger(__name__) # pylint: disable=invalid-name
_snake_case = "\n Examples:\n ```py\n >>> from PIL import Image\n >>> import torch\n >>> from diffusers import DiffusionPipeline\n >>> from diffusers.utils import export_to_gif, load_image\n\n >>> device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n\n >>> repo = \"openai/shap-e-img2img\"\n >>> pipe = DiffusionPipeline.from_pretrained(repo, torch_dtype=torch.float16)\n >>> pipe = pipe.to(device)\n\n >>> guidance_scale = 3.0\n >>> image_url = \"https://hf.co/datasets/diffusers/docs-images/resolve/main/shap-e/corgi.png\"\n >>> image = load_image(image_url).convert(\"RGB\")\n\n >>> images = pipe(\n ... image,\n ... guidance_scale=guidance_scale,\n ... num_inference_steps=64,\n ... frame_size=256,\n ... ).images\n\n >>> gif_path = export_to_gif(images[0], \"corgi_3d.gif\")\n ```\n"
@dataclass
class lowercase ( UpperCamelCase__ ):
_a = 42
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a , _a , _a , ) -> List[Any]:
super().__init__()
self.register_modules(
prior=_a , image_encoder=_a , image_processor=_a , scheduler=_a , renderer=_a , )
def a__ ( self , _a , _a , _a , _a , _a , _a ) -> str:
if latents is None:
_A : str = randn_tensor(_a , generator=_a , device=_a , dtype=_a )
else:
if latents.shape != shape:
raise ValueError(F'''Unexpected latents shape, got {latents.shape}, expected {shape}''' )
_A : Union[str, Any] = latents.to(_a )
_A : int = latents * scheduler.init_noise_sigma
return latents
def a__ ( self , _a=0 ) -> Optional[Any]:
if is_accelerate_available():
from accelerate import cpu_offload
else:
raise ImportError("""Please install accelerate via `pip install accelerate`""" )
_A : str = torch.device(F'''cuda:{gpu_id}''' )
_A : Any = [self.image_encoder, self.prior]
for cpu_offloaded_model in models:
if cpu_offloaded_model is not None:
cpu_offload(_a , _a )
@property
def a__ ( self ) -> List[Any]:
if self.device != torch.device("""meta""" ) or not hasattr(self.image_encoder , """_hf_hook""" ):
return self.device
for module in self.image_encoder.modules():
if (
hasattr(_a , """_hf_hook""" )
and hasattr(module._hf_hook , """execution_device""" )
and module._hf_hook.execution_device is not None
):
return torch.device(module._hf_hook.execution_device )
return self.device
def a__ ( self , _a , _a , _a , _a , ) -> Tuple:
if isinstance(_a , _a ) and isinstance(image[0] , torch.Tensor ):
_A : int = torch.cat(_a , axis=0 ) if image[0].ndim == 4 else torch.stack(_a , axis=0 )
if not isinstance(_a , torch.Tensor ):
_A : Dict = self.image_processor(_a , return_tensors="""pt""" ).pixel_values[0].unsqueeze(0 )
_A : int = image.to(dtype=self.image_encoder.dtype , device=_a )
_A : List[Any] = self.image_encoder(_a )["""last_hidden_state"""]
_A : List[Any] = image_embeds[:, 1:, :].contiguous() # batch_size, dim, 256
_A : Dict = image_embeds.repeat_interleave(_a , dim=0 )
if do_classifier_free_guidance:
_A : str = torch.zeros_like(_a )
# For classifier free guidance, we need to do two forward passes.
# Here we concatenate the unconditional and text embeddings into a single batch
# to avoid doing two forward passes
_A : List[str] = torch.cat([negative_image_embeds, image_embeds] )
return image_embeds
@torch.no_grad()
@replace_example_docstring(_a )
def __call__( self , _a , _a = 1 , _a = 25 , _a = None , _a = None , _a = 4.0 , _a = 64 , _a = "pil" , _a = True , ) -> Union[str, Any]:
if isinstance(_a , PIL.Image.Image ):
_A : List[Any] = 1
elif isinstance(_a , torch.Tensor ):
_A : Any = image.shape[0]
elif isinstance(_a , _a ) and isinstance(image[0] , (torch.Tensor, PIL.Image.Image) ):
_A : Union[str, Any] = len(_a )
else:
raise ValueError(
F'''`image` has to be of type `PIL.Image.Image`, `torch.Tensor`, `List[PIL.Image.Image]` or `List[torch.Tensor]` but is {type(_a )}''' )
_A : Optional[int] = self._execution_device
_A : Tuple = batch_size * num_images_per_prompt
_A : List[Any] = guidance_scale > 1.0
_A : Optional[Any] = self._encode_image(_a , _a , _a , _a )
# prior
self.scheduler.set_timesteps(_a , device=_a )
_A : Optional[int] = self.scheduler.timesteps
_A : List[str] = self.prior.config.num_embeddings
_A : int = self.prior.config.embedding_dim
_A : Optional[Any] = self.prepare_latents(
(batch_size, num_embeddings * embedding_dim) , image_embeds.dtype , _a , _a , _a , self.scheduler , )
# YiYi notes: for testing only to match ldm, we can directly create a latents with desired shape: batch_size, num_embeddings, embedding_dim
_A : List[Any] = latents.reshape(latents.shape[0] , _a , _a )
for i, t in enumerate(self.progress_bar(_a ) ):
# expand the latents if we are doing classifier free guidance
_A : List[str] = torch.cat([latents] * 2 ) if do_classifier_free_guidance else latents
_A : int = self.scheduler.scale_model_input(_a , _a )
_A : Tuple = self.prior(
_a , timestep=_a , proj_embedding=_a , ).predicted_image_embedding
# remove the variance
_A , _A : Optional[Any] = noise_pred.split(
scaled_model_input.shape[2] , dim=2 ) # batch_size, num_embeddings, embedding_dim
if do_classifier_free_guidance is not None:
_A , _A : Dict = noise_pred.chunk(2 )
_A : Tuple = noise_pred_uncond + guidance_scale * (noise_pred - noise_pred_uncond)
_A : int = self.scheduler.step(
_a , timestep=_a , sample=_a , ).prev_sample
if output_type == "latent":
return ShapEPipelineOutput(images=_a )
_A : List[str] = []
for i, latent in enumerate(_a ):
print()
_A : List[str] = self.renderer.decode(
latent[None, :] , _a , size=_a , ray_batch_size=4096 , n_coarse_samples=64 , n_fine_samples=128 , )
images.append(_a )
_A : List[Any] = torch.stack(_a )
if output_type not in ["np", "pil"]:
raise ValueError(F'''Only the output types `pil` and `np` are supported not output_type={output_type}''' )
_A : List[str] = images.cpu().numpy()
if output_type == "pil":
_A : List[Any] = [self.numpy_to_pil(_a ) for image in images]
# Offload last model to CPU
if hasattr(self , """final_offload_hook""" ) and self.final_offload_hook is not None:
self.final_offload_hook.offload()
if not return_dict:
return (images,)
return ShapEPipelineOutput(images=_a )
| 26 | 1 |
import argparse
import shutil
import time
from json import JSONDecodeError
from logging import getLogger
from pathlib import Path
from typing import Dict, List
import torch
from torch.utils.data import DataLoader
from tqdm import tqdm
from transformers import AutoModelForSeqaSeqLM, AutoTokenizer
from utils import (
SeqaSeqDataset,
calculate_bleu,
calculate_rouge,
chunks,
lmap,
load_json,
parse_numeric_n_bool_cl_kwargs,
save_json,
use_task_specific_params,
write_txt_file,
)
_snake_case = getLogger(__name__)
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ = 8,snake_case_ = 1024,snake_case_="val",snake_case_=None,snake_case_=False,snake_case_="summarization",snake_case_=None,snake_case_=1,snake_case_ = None,snake_case_="",**snake_case_,):
_A : Dict = str(snake_case_ )
assert local_rank is not None
torch.distributed.init_process_group(backend="""nccl""",rank=snake_case_ )
_A : Tuple = Path(snake_case_ )
_A : List[Any] = save_dir.joinpath(f'''rank_{local_rank}_output.json''' )
torch.cuda.set_device(snake_case_ )
_A : Union[str, Any] = AutoModelForSeqaSeqLM.from_pretrained(snake_case_ ).cuda()
if fpaa:
_A : Any = model.half()
# determine if we need to increase num_beams
use_task_specific_params(snake_case_,snake_case_ ) # update config with task specific params
_A : str = generate_kwargs.pop("""num_beams""",model.config.num_beams ) # AttributeError risk?
if num_return_sequences > num_beams:
_A : int = num_return_sequences
_A : Optional[Any] = AutoTokenizer.from_pretrained(snake_case_ )
logger.info(f'''Inferred tokenizer type: {tokenizer.__class__}''' ) # if this is wrong, check config.model_type.
if max_source_length is None:
_A : Optional[int] = tokenizer.model_max_length
if prefix is None:
_A : Tuple = prefix or getattr(model.config,"""prefix""","""""" ) or """"""
_A : Optional[int] = SeqaSeqDataset(
snake_case_,snake_case_,snake_case_,max_target_length=1024,type_path=snake_case_,n_obs=snake_case_,prefix=snake_case_,**snake_case_,)
# I set shuffle=True for a more accurate progress bar.
# If all the longest samples are first, the prog bar estimate is too high at the beginning.
_A : Optional[int] = ds.make_sortish_sampler(snake_case_,distributed=snake_case_,add_extra_examples=snake_case_,shuffle=snake_case_ )
_A : Dict = DataLoader(snake_case_,sampler=snake_case_,batch_size=snake_case_,collate_fn=ds.collate_fn )
_A : Optional[Any] = []
for batch in tqdm(snake_case_ ):
_A : Tuple = model.generate(
input_ids=batch["""input_ids"""].to(model.device ),attention_mask=batch["""attention_mask"""].to(model.device ),num_return_sequences=snake_case_,num_beams=snake_case_,**snake_case_,)
_A : Any = tokenizer.batch_decode(snake_case_,skip_special_tokens=snake_case_,clean_up_tokenization_spaces=snake_case_ )
_A : Dict = batch["""ids"""]
if num_return_sequences > 1:
_A : Any = chunks(snake_case_,snake_case_ ) # batch size chunks, each of size num_return_seq
for i, pred in enumerate(snake_case_ ):
results.append({"""pred""": pred, """id""": ids[i].item()} )
save_json(snake_case_,snake_case_ )
return results, sampler.num_replicas
def lowerCAmelCase_ ( ):
_A : Tuple = argparse.ArgumentParser(
epilog="""Unspecified args like --num_beams=2 --decoder_start_token_id=4 are passed to model.generate""" )
parser.add_argument("""--data_dir""",type=snake_case_,help="""like cnn_dm/test.source""" )
parser.add_argument(
"""--model_name""",type=snake_case_,help="""like facebook/bart-large-cnn,t5-base, etc.""",default="""sshleifer/distilbart-xsum-12-3""",)
parser.add_argument("""--save_dir""",type=snake_case_,help="""where to save""",default="""tmp_gen""" )
parser.add_argument("""--max_source_length""",type=snake_case_,default=snake_case_ )
parser.add_argument(
"""--type_path""",type=snake_case_,default="""test""",help="""which subset to evaluate typically train/val/test""" )
parser.add_argument("""--task""",type=snake_case_,default="""summarization""",help="""used for task_specific_params + metrics""" )
parser.add_argument("""--bs""",type=snake_case_,default=8,required=snake_case_,help="""batch size""" )
parser.add_argument(
"""--local_rank""",type=snake_case_,default=-1,required=snake_case_,help="""should be passed by distributed.launch""" )
parser.add_argument(
"""--n_obs""",type=snake_case_,default=snake_case_,required=snake_case_,help="""How many observations. Defaults to all.""" )
parser.add_argument(
"""--num_return_sequences""",type=snake_case_,default=1,required=snake_case_,help="""How many sequences to return""" )
parser.add_argument(
"""--sync_timeout""",type=snake_case_,default=600,required=snake_case_,help="""How long should master process wait for other processes to finish.""",)
parser.add_argument("""--src_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument("""--tgt_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument(
"""--prefix""",type=snake_case_,required=snake_case_,default=snake_case_,help="""will be added to the begininng of src examples""" )
parser.add_argument("""--fp16""",action="""store_true""" )
parser.add_argument("""--debug""",action="""store_true""" )
_A : Union[str, Any] = time.time()
_A , _A : List[str] = parser.parse_known_args()
_A : List[str] = parse_numeric_n_bool_cl_kwargs(snake_case_ )
if generate_kwargs and args.local_rank <= 0:
print(f'''parsed the following generate kwargs: {generate_kwargs}''' )
_A : Dict = Path(args.save_dir + """_tmp""" )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ ) # this handles locking.
_A : int = list(json_save_dir.glob("""rank_*.json""" ) )
if intermediate_files:
raise ValueError(f'''Found files at {json_save_dir} please move or remove them.''' )
# In theory, a node could finish and save before another node hits this. If this happens, we can address later.
_A : Any = {}
if args.src_lang is not None:
_A : int = args.src_lang
if args.tgt_lang is not None:
_A : Dict = args.tgt_lang
Path(args.save_dir ).mkdir(exist_ok=snake_case_ )
_A , _A : str = eval_data_dir(
args.data_dir,snake_case_,args.model_name,type_path=args.type_path,bs=args.bs,fpaa=args.fpaa,task=args.task,local_rank=args.local_rank,n_obs=args.n_obs,max_source_length=args.max_source_length,num_return_sequences=args.num_return_sequences,prefix=args.prefix,dataset_kwargs=snake_case_,**snake_case_,)
if args.local_rank <= 0:
_A : List[Any] = Path(args.save_dir )
save_dir.mkdir(exist_ok=snake_case_ )
_A : Tuple = gather_results_from_each_node(snake_case_,snake_case_,args.sync_timeout )
_A : Optional[int] = combine_partial_results(snake_case_ )
if args.num_return_sequences > 1:
_A : Optional[Any] = save_dir.joinpath("""pseudolabel_results.json""" )
print(f'''Saving aggregated results at {save_path}, intermediate in {json_save_dir}/''' )
save_json(snake_case_,snake_case_ )
return
_A : List[str] = Path(args.data_dir ).joinpath(args.type_path + """.target""" )
with open(snake_case_ ) as f:
_A : int = [x.rstrip() for x in f.readlines()][: len(snake_case_ )]
# Calculate metrics, save metrics, and save _generations.txt
_A : Dict = """translation""" in args.task
_A : Optional[Any] = calculate_bleu if calc_bleu else calculate_rouge
_A : Tuple = """bleu""" if calc_bleu else """rouge"""
_A : Dict = score_fn(snake_case_,snake_case_ )
_A : List[Any] = len(snake_case_ )
_A : Optional[int] = time.time() - start_time
_A : Dict = round(runtime / metrics["""n_obs"""],4 )
_A : Dict = num_replicas
# TODO(@stas00): add whatever metadata to metrics
_A : Any = save_dir.joinpath(f'''{args.type_path}_{metric_name}.json''' )
save_json(snake_case_,snake_case_,indent=snake_case_ )
print(snake_case_ )
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}_generations.txt''' ) )
if args.debug:
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}.target''' ) )
else:
shutil.rmtree(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
_A : Dict = []
for partial_result in partial_results:
records.extend(snake_case_ )
_A : Optional[Any] = sorted(snake_case_,key=lambda snake_case_ : x["id"] )
_A : List[str] = [x["""pred"""] for x in records]
return preds
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
# WAIT FOR lots of .json files
_A : Optional[Any] = time.time()
logger.info("""waiting for all nodes to finish""" )
_A : List[str] = None
while (time.time() - start_wait) < timeout:
_A : str = list(save_dir.glob("""rank_*.json""" ) )
if len(snake_case_ ) < num_replicas:
continue
try:
# make sure all json files are fully saved
_A : List[str] = lmap(snake_case_,snake_case_ )
return json_data
except JSONDecodeError:
continue
else:
raise TimeoutError("""Rank 0 gave up on waiting for other processes""" )
# Unreachable
if __name__ == "__main__":
# Usage for MT:
run_generate()
| 26 |
import argparse
import collections
import json
from pathlib import Path
import requests
import torch
import yaml
from huggingface_hub import hf_hub_download
from PIL import Image
from transformers import (
MobileViTImageProcessor,
MobileViTVaConfig,
MobileViTVaForImageClassification,
MobileViTVaForSemanticSegmentation,
)
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
def lowerCAmelCase_ ( snake_case_ ):
print("""Loading config file...""" )
def flatten_yaml_as_dict(snake_case_,snake_case_="",snake_case_="." ):
_A : Union[str, Any] = []
for k, v in d.items():
_A : Optional[int] = parent_key + sep + k if parent_key else k
if isinstance(snake_case_,collections.abc.MutableMapping ):
items.extend(flatten_yaml_as_dict(snake_case_,snake_case_,sep=snake_case_ ).items() )
else:
items.append((new_key, v) )
return dict(snake_case_ )
_A : List[Any] = argparse.Namespace()
with open(snake_case_,"""r""" ) as yaml_file:
try:
_A : List[Any] = yaml.load(snake_case_,Loader=yaml.FullLoader )
_A : Optional[int] = flatten_yaml_as_dict(snake_case_ )
for k, v in flat_cfg.items():
setattr(snake_case_,snake_case_,snake_case_ )
except yaml.YAMLError as exc:
logger.error("""Error while loading config file: {}. Error message: {}""".format(snake_case_,str(snake_case_ ) ) )
return config
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Optional[Any] = MobileViTVaConfig()
_A : Tuple = False
# dataset
if task_name.startswith("""imagenet1k_""" ):
_A : Dict = 1000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : int = 384
else:
_A : int = 256
_A : List[str] = """imagenet-1k-id2label.json"""
elif task_name.startswith("""imagenet21k_to_1k_""" ):
_A : Union[str, Any] = 21000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : str = 384
else:
_A : List[Any] = 256
_A : List[str] = """imagenet-22k-id2label.json"""
elif task_name.startswith("""ade20k_""" ):
_A : int = 151
_A : int = 512
_A : Optional[int] = """ade20k-id2label.json"""
_A : Any = True
elif task_name.startswith("""voc_""" ):
_A : List[Any] = 21
_A : Dict = 512
_A : Dict = """pascal-voc-id2label.json"""
_A : int = True
# orig_config
_A : Any = load_orig_config_file(snake_case_ )
assert getattr(snake_case_,"""model.classification.name""",-1 ) == "mobilevit_v2", "Invalid model"
_A : List[Any] = getattr(snake_case_,"""model.classification.mitv2.width_multiplier""",1.0 )
assert (
getattr(snake_case_,"""model.classification.mitv2.attn_norm_layer""",-1 ) == "layer_norm_2d"
), "Norm layers other than layer_norm_2d is not supported"
_A : str = getattr(snake_case_,"""model.classification.activation.name""","""swish""" )
# config.image_size == getattr(orig_config, 'sampler.bs.crop_size_width', 256)
if is_segmentation_model:
_A : Optional[int] = getattr(snake_case_,"""model.segmentation.output_stride""",16 )
if "_deeplabv3" in task_name:
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_rates""",[12, 24, 36] )
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_out_channels""",512 )
_A : str = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_dropout""",0.1 )
# id2label
_A : List[Any] = """huggingface/label-files"""
_A : List[Any] = json.load(open(hf_hub_download(snake_case_,snake_case_,repo_type="""dataset""" ),"""r""" ) )
_A : str = {int(snake_case_ ): v for k, v in idalabel.items()}
_A : str = idalabel
_A : Dict = {v: k for k, v in idalabel.items()}
return config
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Any = dct.pop(snake_case_ )
_A : Union[str, Any] = val
def lowerCAmelCase_ ( snake_case_,snake_case_=False ):
if base_model:
_A : Optional[int] = """"""
else:
_A : Dict = """mobilevitv2."""
_A : int = []
for k in state_dict.keys():
if k[:8] == "encoder.":
_A : Any = k[8:]
else:
_A : List[str] = k
if ".block." in k:
_A : Any = k_new.replace(""".block.""",""".""" )
if ".conv." in k:
_A : List[Any] = k_new.replace(""".conv.""",""".convolution.""" )
if ".norm." in k:
_A : Any = k_new.replace(""".norm.""",""".normalization.""" )
if "conv_1." in k:
_A : int = k_new.replace("""conv_1.""",f'''{model_prefix}conv_stem.''' )
for i in [1, 2]:
if f'''layer_{i}.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.''',f'''{model_prefix}encoder.layer.{i-1}.layer.''' )
if ".exp_1x1." in k:
_A : Tuple = k_new.replace(""".exp_1x1.""",""".expand_1x1.""" )
if ".red_1x1." in k:
_A : Optional[int] = k_new.replace(""".red_1x1.""",""".reduce_1x1.""" )
for i in [3, 4, 5]:
if f'''layer_{i}.0.''' in k:
_A : Optional[int] = k_new.replace(f'''layer_{i}.0.''',f'''{model_prefix}encoder.layer.{i-1}.downsampling_layer.''' )
if f'''layer_{i}.1.local_rep.0.''' in k:
_A : Union[str, Any] = k_new.replace(f'''layer_{i}.1.local_rep.0.''',f'''{model_prefix}encoder.layer.{i-1}.conv_kxk.''' )
if f'''layer_{i}.1.local_rep.1.''' in k:
_A : str = k_new.replace(f'''layer_{i}.1.local_rep.1.''',f'''{model_prefix}encoder.layer.{i-1}.conv_1x1.''' )
for i in [3, 4, 5]:
if i == 3:
_A : Optional[int] = [0, 1]
elif i == 4:
_A : Union[str, Any] = [0, 1, 2, 3]
elif i == 5:
_A : Optional[Any] = [0, 1, 2]
for j in j_in:
if f'''layer_{i}.1.global_rep.{j}.''' in k:
_A : Union[str, Any] = k_new.replace(
f'''layer_{i}.1.global_rep.{j}.''',f'''{model_prefix}encoder.layer.{i-1}.transformer.layer.{j}.''' )
if f'''layer_{i}.1.global_rep.{j+1}.''' in k:
_A : List[str] = k_new.replace(
f'''layer_{i}.1.global_rep.{j+1}.''',f'''{model_prefix}encoder.layer.{i-1}.layernorm.''' )
if f'''layer_{i}.1.conv_proj.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.1.conv_proj.''',f'''{model_prefix}encoder.layer.{i-1}.conv_projection.''' )
if "pre_norm_attn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_attn.0.""","""layernorm_before.""" )
if "pre_norm_attn.1." in k:
_A : str = k_new.replace("""pre_norm_attn.1.""","""attention.""" )
if "pre_norm_ffn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_ffn.0.""","""layernorm_after.""" )
if "pre_norm_ffn.1." in k:
_A : Dict = k_new.replace("""pre_norm_ffn.1.""","""ffn.conv1.""" )
if "pre_norm_ffn.3." in k:
_A : List[str] = k_new.replace("""pre_norm_ffn.3.""","""ffn.conv2.""" )
if "classifier.1." in k:
_A : List[str] = k_new.replace("""classifier.1.""","""classifier.""" )
if "seg_head." in k:
_A : List[Any] = k_new.replace("""seg_head.""","""segmentation_head.""" )
if ".aspp_layer." in k:
_A : List[Any] = k_new.replace(""".aspp_layer.""",""".""" )
if ".aspp_pool." in k:
_A : Optional[Any] = k_new.replace(""".aspp_pool.""",""".""" )
rename_keys.append((k, k_new) )
return rename_keys
def lowerCAmelCase_ ( snake_case_ ):
_A : Tuple = []
for k in state_dict.keys():
if k.startswith("""seg_head.aux_head.""" ):
keys_to_ignore.append(snake_case_ )
for k in keys_to_ignore:
state_dict.pop(snake_case_,snake_case_ )
def lowerCAmelCase_ ( ):
_A : Dict = """http://images.cocodataset.org/val2017/000000039769.jpg"""
# url = "https://cdn.britannica.com/86/141086-050-9D7C75EE/Gulfstream-G450-business-jet-passengers.jpg"
_A : List[Any] = Image.open(requests.get(snake_case_,stream=snake_case_ ).raw )
return im
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : List[Any] = get_mobilevitva_config(snake_case_,snake_case_ )
# load original state_dict
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
# load huggingface model
if task_name.startswith("""ade20k_""" ) or task_name.startswith("""voc_""" ):
_A : Optional[Any] = MobileViTVaForSemanticSegmentation(snake_case_ ).eval()
_A : str = False
else:
_A : int = MobileViTVaForImageClassification(snake_case_ ).eval()
_A : List[Any] = False
# remove and rename some keys of load the original model
_A : List[Any] = checkpoint
remove_unused_keys(snake_case_ )
_A : Optional[Any] = create_rename_keys(snake_case_,base_model=snake_case_ )
for rename_key_src, rename_key_dest in rename_keys:
rename_key(snake_case_,snake_case_,snake_case_ )
# load modified state_dict
model.load_state_dict(snake_case_ )
# Check outputs on an image, prepared by MobileViTImageProcessor
_A : str = MobileViTImageProcessor(crop_size=config.image_size,size=config.image_size + 32 )
_A : List[Any] = image_processor(images=prepare_img(),return_tensors="""pt""" )
_A : Optional[Any] = model(**snake_case_ )
# verify classification model
if task_name.startswith("""imagenet""" ):
_A : List[Any] = outputs.logits
_A : Optional[int] = logits.argmax(-1 ).item()
print("""Predicted class:""",model.config.idalabel[predicted_class_idx] )
if task_name.startswith("""imagenet1k_256""" ) and config.width_multiplier == 1.0:
# expected_logits for base variant
_A : int = torch.tensor([-1.63_36e00, -7.32_04e-02, -5.18_83e-01] )
assert torch.allclose(logits[0, :3],snake_case_,atol=1e-4 )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ )
print(f'''Saving model {task_name} to {pytorch_dump_folder_path}''' )
model.save_pretrained(snake_case_ )
print(f'''Saving image processor to {pytorch_dump_folder_path}''' )
image_processor.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
"--task",
default="imagenet1k_256",
type=str,
help=(
"Name of the task for which the MobileViTV2 model you'd like to convert is trained on . "
"\n Classification (ImageNet-1k)\n - MobileViTV2 (256x256) : imagenet1k_256\n - MobileViTV2 (Trained on 256x256 and Finetuned on 384x384) : imagenet1k_384\n - MobileViTV2 (Trained on ImageNet-21k and Finetuned on ImageNet-1k 256x256) :\n imagenet21k_to_1k_256\n - MobileViTV2 (Trained on ImageNet-21k, Finetuned on ImageNet-1k 256x256, and Finetuned on\n ImageNet-1k 384x384) : imagenet21k_to_1k_384\n Segmentation\n - ADE20K Dataset : ade20k_deeplabv3\n - Pascal VOC 2012 Dataset: voc_deeplabv3\n "
),
choices=[
"imagenet1k_256",
"imagenet1k_384",
"imagenet21k_to_1k_256",
"imagenet21k_to_1k_384",
"ade20k_deeplabv3",
"voc_deeplabv3",
],
)
parser.add_argument(
"--orig_checkpoint_path", required=True, type=str, help="Path to the original state dict (.pt file)."
)
parser.add_argument("--orig_config_path", required=True, type=str, help="Path to the original config file.")
parser.add_argument(
"--pytorch_dump_folder_path", required=True, type=str, help="Path to the output PyTorch model directory."
)
_snake_case = parser.parse_args()
convert_mobilevitva_checkpoint(
args.task, args.orig_checkpoint_path, args.orig_config_path, args.pytorch_dump_folder_path
)
| 26 | 1 |
from __future__ import annotations
import math
def lowerCAmelCase_ ( snake_case_ ):
if num <= 0:
_A : int = f'''{num}: Invalid input, please enter a positive integer.'''
raise ValueError(snake_case_ )
_A : Optional[Any] = [True] * (num + 1)
_A : Optional[Any] = []
_A : Optional[Any] = 2
_A : Tuple = int(math.sqrt(snake_case_ ) )
while start <= end:
# If start is a prime
if sieve[start] is True:
prime.append(snake_case_ )
# Set multiples of start be False
for i in range(start * start,num + 1,snake_case_ ):
if sieve[i] is True:
_A : Union[str, Any] = False
start += 1
for j in range(end + 1,num + 1 ):
if sieve[j] is True:
prime.append(snake_case_ )
return prime
if __name__ == "__main__":
print(prime_sieve(int(input("Enter a positive integer: ").strip())))
| 26 |
import torch
from diffusers import DPMSolverSDEScheduler
from diffusers.utils import torch_device
from diffusers.utils.testing_utils import require_torchsde
from .test_schedulers import SchedulerCommonTest
@require_torchsde
class lowercase ( UpperCamelCase__ ):
_a = (DPMSolverSDEScheduler,)
_a = 1_0
def a__ ( self , **_a ) -> Optional[Any]:
_A : str = {
"""num_train_timesteps""": 1100,
"""beta_start""": 0.0001,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
"""noise_sampler_seed""": 0,
}
config.update(**_a )
return config
def a__ ( self ) -> Tuple:
for timesteps in [10, 50, 100, 1000]:
self.check_over_configs(num_train_timesteps=_a )
def a__ ( self ) -> Optional[int]:
for beta_start, beta_end in zip([0.00001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ):
self.check_over_configs(beta_start=_a , beta_end=_a )
def a__ ( self ) -> Any:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=_a )
def a__ ( self ) -> Optional[int]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=_a )
def a__ ( self ) -> Optional[int]:
_A : Any = self.scheduler_classes[0]
_A : List[str] = self.get_scheduler_config()
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Dict = self.dummy_model()
_A : Any = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Dict = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : str = model(_a , _a )
_A : List[Any] = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Dict = torch.sum(torch.abs(_a ) )
_A : Dict = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.47821044921875 ) < 1e-2
assert abs(result_mean.item() - 0.2178705964565277 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59352111816406 ) < 1e-2
assert abs(result_mean.item() - 0.22342906892299652 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Optional[Any]:
_A : Dict = self.scheduler_classes[0]
_A : Optional[int] = self.get_scheduler_config(prediction_type="""v_prediction""" )
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Tuple = self.dummy_model()
_A : int = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Tuple = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : int = scheduler.scale_model_input(_a , _a )
_A : Tuple = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Optional[Any] = torch.sum(torch.abs(_a ) )
_A : List[Any] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 124.77149200439453 ) < 1e-2
assert abs(result_mean.item() - 0.16226289014816284 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 128.1663360595703 ) < 1e-2
assert abs(result_mean.item() - 0.16688326001167297 ) < 1e-3
else:
assert abs(result_sum.item() - 119.8487548828125 ) < 1e-2
assert abs(result_mean.item() - 0.1560530662536621 ) < 1e-3
def a__ ( self ) -> List[str]:
_A : Union[str, Any] = self.scheduler_classes[0]
_A : List[Any] = self.get_scheduler_config()
_A : List[str] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Union[str, Any] = self.dummy_model()
_A : Optional[Any] = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
_A : int = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Dict = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : str = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.46957397460938 ) < 1e-2
assert abs(result_mean.item() - 0.21805934607982635 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59353637695312 ) < 1e-2
assert abs(result_mean.item() - 0.22342908382415771 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Union[str, Any]:
_A : List[Any] = self.scheduler_classes[0]
_A : Optional[Any] = self.get_scheduler_config()
_A : int = scheduler_class(**_a , use_karras_sigmas=_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Optional[Any] = self.dummy_model()
_A : Dict = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
_A : str = sample.to(_a )
for t in scheduler.timesteps:
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : List[str] = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : List[str] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 176.66974135742188 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 177.63653564453125 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
else:
assert abs(result_sum.item() - 170.3135223388672 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
| 26 | 1 |
from math import asin, atan, cos, radians, sin, sqrt, tan
_snake_case = 6_3_7_8_1_3_7.0
_snake_case = 6_3_5_6_7_5_2.3_1_4_2_4_5
_snake_case = 6378137
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : Any = (AXIS_A - AXIS_B) / AXIS_A
_A : Optional[int] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : List[str] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : Optional[Any] = radians(snake_case_ )
_A : str = radians(snake_case_ )
# Equation
_A : Dict = sin((phi_a - phi_a) / 2 )
_A : List[str] = sin((lambda_a - lambda_a) / 2 )
# Square both values
sin_sq_phi *= sin_sq_phi
sin_sq_lambda *= sin_sq_lambda
_A : Optional[int] = sqrt(sin_sq_phi + (cos(snake_case_ ) * cos(snake_case_ ) * sin_sq_lambda) )
return 2 * RADIUS * asin(snake_case_ )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 |
# DISCLAIMER: This file is strongly influenced by https://github.com/yang-song/score_sde_pytorch
import math
from typing import Union
import torch
from ..configuration_utils import ConfigMixin, register_to_config
from ..utils import randn_tensor
from .scheduling_utils import SchedulerMixin
class lowercase ( UpperCamelCase__,UpperCamelCase__ ):
_a = 1
@register_to_config
def __init__( self , _a=2000 , _a=0.1 , _a=20 , _a=1e-3 ) -> List[Any]:
_A : Dict = None
_A : List[Any] = None
_A : Dict = None
def a__ ( self , _a , _a = None ) -> Union[str, Any]:
_A : Union[str, Any] = torch.linspace(1 , self.config.sampling_eps , _a , device=_a )
def a__ ( self , _a , _a , _a , _a=None ) -> Dict:
if self.timesteps is None:
raise ValueError(
"""`self.timesteps` is not set, you need to run 'set_timesteps' after creating the scheduler""" )
# TODO(Patrick) better comments + non-PyTorch
# postprocess model score
_A : Any = (
-0.25 * t**2 * (self.config.beta_max - self.config.beta_min) - 0.5 * t * self.config.beta_min
)
_A : List[Any] = torch.sqrt(1.0 - torch.exp(2.0 * log_mean_coeff ) )
_A : List[str] = std.flatten()
while len(std.shape ) < len(score.shape ):
_A : List[Any] = std.unsqueeze(-1 )
_A : int = -score / std
# compute
_A : Tuple = -1.0 / len(self.timesteps )
_A : str = self.config.beta_min + t * (self.config.beta_max - self.config.beta_min)
_A : List[str] = beta_t.flatten()
while len(beta_t.shape ) < len(x.shape ):
_A : Union[str, Any] = beta_t.unsqueeze(-1 )
_A : Tuple = -0.5 * beta_t * x
_A : Tuple = torch.sqrt(_a )
_A : Dict = drift - diffusion**2 * score
_A : Dict = x + drift * dt
# add noise
_A : Any = randn_tensor(x.shape , layout=x.layout , generator=_a , device=x.device , dtype=x.dtype )
_A : str = x_mean + diffusion * math.sqrt(-dt ) * noise
return x, x_mean
def __len__( self ) -> Optional[Any]:
return self.config.num_train_timesteps
| 26 | 1 |
def lowerCAmelCase_ ( snake_case_ ):
if any(not isinstance(snake_case_,snake_case_ ) or x < 0 for x in sequence ):
raise TypeError("""Sequence must be list of non-negative integers""" )
for _ in range(len(snake_case_ ) ):
for i, (rod_upper, rod_lower) in enumerate(zip(snake_case_,sequence[1:] ) ):
if rod_upper > rod_lower:
sequence[i] -= rod_upper - rod_lower
sequence[i + 1] += rod_upper - rod_lower
return sequence
if __name__ == "__main__":
assert bead_sort([5, 4, 3, 2, 1]) == [1, 2, 3, 4, 5]
assert bead_sort([7, 9, 4, 3, 5]) == [3, 4, 5, 7, 9]
| 26 |
import os
from shutil import copyfile
from typing import List, Optional, Tuple
from ...tokenization_utils import AddedToken
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import is_sentencepiece_available, logging
if is_sentencepiece_available():
from .tokenization_fnet import FNetTokenizer
else:
_snake_case = None
_snake_case = logging.get_logger(__name__)
_snake_case = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.json"}
_snake_case = {
"vocab_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/spiece.model",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/spiece.model",
},
"tokenizer_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/tokenizer.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/tokenizer.json",
},
}
_snake_case = {
"google/fnet-base": 512,
"google/fnet-large": 512,
}
_snake_case = "▁"
class lowercase ( UpperCamelCase__ ):
_a = VOCAB_FILES_NAMES
_a = PRETRAINED_VOCAB_FILES_MAP
_a = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
_a = ["input_ids", "token_type_ids"]
_a = FNetTokenizer
def __init__( self , _a=None , _a=None , _a=False , _a=True , _a=True , _a="<unk>" , _a="[SEP]" , _a="<pad>" , _a="[CLS]" , _a="[MASK]" , **_a , ) -> Optional[int]:
# Mask token behave like a normal word, i.e. include the space before it and
# is included in the raw text, there should be a match in a non-normalized sentence.
_A : int = (
AddedToken(_a , lstrip=_a , rstrip=_a , normalized=_a )
if isinstance(_a , _a )
else mask_token
)
super().__init__(
_a , tokenizer_file=_a , do_lower_case=_a , remove_space=_a , keep_accents=_a , unk_token=_a , sep_token=_a , pad_token=_a , cls_token=_a , mask_token=_a , **_a , )
_A : Optional[int] = do_lower_case
_A : List[Any] = remove_space
_A : str = keep_accents
_A : int = vocab_file
_A : int = False if not self.vocab_file else True
def a__ ( self , _a , _a = None ) -> List[int]:
_A : str = [self.sep_token_id]
_A : Dict = [self.cls_token_id]
if token_ids_a is None:
return cls + token_ids_a + sep
return cls + token_ids_a + sep + token_ids_a + sep
def a__ ( self , _a , _a = None ) -> List[int]:
_A : Any = [self.sep_token_id]
_A : Any = [self.cls_token_id]
if token_ids_a is None:
return len(cls + token_ids_a + sep ) * [0]
return len(cls + token_ids_a + sep ) * [0] + len(token_ids_a + sep ) * [1]
def a__ ( self , _a , _a = None ) -> Tuple[str]:
if not os.path.isdir(_a ):
logger.error(F'''Vocabulary path ({save_directory}) should be a directory''' )
return
_A : List[str] = os.path.join(
_a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
if os.path.abspath(self.vocab_file ) != os.path.abspath(_a ):
copyfile(self.vocab_file , _a )
return (out_vocab_file,)
| 26 | 1 |
from typing import Callable, Dict, Optional, Tuple
import torch
from torch import nn
from torch.distributions import (
AffineTransform,
Distribution,
Independent,
NegativeBinomial,
Normal,
StudentT,
TransformedDistribution,
)
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a=None , _a=None , _a=0 ) -> List[Any]:
_A : int = 1.0 if scale is None else scale
_A : Dict = 0.0 if loc is None else loc
super().__init__(_a , [AffineTransform(loc=self.loc , scale=self.scale , event_dim=_a )] )
@property
def a__ ( self ) -> Tuple:
return self.base_dist.mean * self.scale + self.loc
@property
def a__ ( self ) -> int:
return self.base_dist.variance * self.scale**2
@property
def a__ ( self ) -> Tuple:
return self.variance.sqrt()
class lowercase ( nn.Module ):
def __init__( self , _a , _a , _a , **_a ) -> None:
super().__init__(**_a )
_A : List[Any] = args_dim
_A : Optional[Any] = nn.ModuleList([nn.Linear(_a , _a ) for dim in args_dim.values()] )
_A : Dict = domain_map
def a__ ( self , _a ) -> Tuple[torch.Tensor]:
_A : Optional[Any] = [proj(_a ) for proj in self.proj]
return self.domain_map(*_a )
class lowercase ( nn.Module ):
def __init__( self , _a ) -> List[Any]:
super().__init__()
_A : Any = function
def a__ ( self , _a , *_a ) -> Optional[int]:
return self.function(_a , *_a )
class lowercase :
_a = 42
_a = 42
_a = 42
def __init__( self , _a = 1 ) -> None:
_A : Tuple = dim
_A : Tuple = {k: dim * self.args_dim[k] for k in self.args_dim}
def a__ ( self , _a ) -> Optional[int]:
if self.dim == 1:
return self.distribution_class(*_a )
else:
return Independent(self.distribution_class(*_a ) , 1 )
def a__ ( self , _a , _a = None , _a = None , ) -> Distribution:
_A : Union[str, Any] = self._base_distribution(_a )
if loc is None and scale is None:
return distr
else:
return AffineTransformed(_a , loc=_a , scale=_a , event_dim=self.event_dim )
@property
def a__ ( self ) -> Tuple:
return () if self.dim == 1 else (self.dim,)
@property
def a__ ( self ) -> int:
return len(self.event_shape )
@property
def a__ ( self ) -> float:
return 0.0
def a__ ( self , _a ) -> nn.Module:
return ParameterProjection(
in_features=_a , args_dim=self.args_dim , domain_map=LambdaLayer(self.domain_map ) , )
def a__ ( self , *_a ) -> Dict:
raise NotImplementedError()
@staticmethod
def a__ ( _a ) -> torch.Tensor:
return (x + torch.sqrt(torch.square(_a ) + 4.0 )) / 2.0
class lowercase ( UpperCamelCase__ ):
_a = {"df": 1, "loc": 1, "scale": 1}
_a = StudentT
@classmethod
def a__ ( cls , _a , _a , _a ) -> int:
_A : int = cls.squareplus(_a ).clamp_min(torch.finfo(scale.dtype ).eps )
_A : Any = 2.0 + cls.squareplus(_a )
return df.squeeze(-1 ), loc.squeeze(-1 ), scale.squeeze(-1 )
class lowercase ( UpperCamelCase__ ):
_a = {"loc": 1, "scale": 1}
_a = Normal
@classmethod
def a__ ( cls , _a , _a ) -> List[Any]:
_A : List[str] = cls.squareplus(_a ).clamp_min(torch.finfo(scale.dtype ).eps )
return loc.squeeze(-1 ), scale.squeeze(-1 )
class lowercase ( UpperCamelCase__ ):
_a = {"total_count": 1, "logits": 1}
_a = NegativeBinomial
@classmethod
def a__ ( cls , _a , _a ) -> Union[str, Any]:
_A : List[str] = cls.squareplus(_a )
return total_count.squeeze(-1 ), logits.squeeze(-1 )
def a__ ( self , _a ) -> Distribution:
_A , _A : Tuple = distr_args
if self.dim == 1:
return self.distribution_class(total_count=_a , logits=_a )
else:
return Independent(self.distribution_class(total_count=_a , logits=_a ) , 1 )
def a__ ( self , _a , _a = None , _a = None ) -> Distribution:
_A , _A : List[Any] = distr_args
if scale is not None:
# See scaling property of Gamma.
logits += scale.log()
return self._base_distribution((total_count, logits) )
| 26 |
from math import asin, atan, cos, radians, sin, sqrt, tan
_snake_case = 6_3_7_8_1_3_7.0
_snake_case = 6_3_5_6_7_5_2.3_1_4_2_4_5
_snake_case = 6378137
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : Any = (AXIS_A - AXIS_B) / AXIS_A
_A : Optional[int] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : List[str] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : Optional[Any] = radians(snake_case_ )
_A : str = radians(snake_case_ )
# Equation
_A : Dict = sin((phi_a - phi_a) / 2 )
_A : List[str] = sin((lambda_a - lambda_a) / 2 )
# Square both values
sin_sq_phi *= sin_sq_phi
sin_sq_lambda *= sin_sq_lambda
_A : Optional[int] = sqrt(sin_sq_phi + (cos(snake_case_ ) * cos(snake_case_ ) * sin_sq_lambda) )
return 2 * RADIUS * asin(snake_case_ )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
import argparse
import io
import requests
import torch
from omegaconf import OmegaConf
from diffusers import AutoencoderKL
from diffusers.pipelines.stable_diffusion.convert_from_ckpt import (
assign_to_checkpoint,
conv_attn_to_linear,
create_vae_diffusers_config,
renew_vae_attention_paths,
renew_vae_resnet_paths,
)
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : List[Any] = checkpoint
_A : Optional[Any] = {}
_A : str = vae_state_dict["""encoder.conv_in.weight"""]
_A : int = vae_state_dict["""encoder.conv_in.bias"""]
_A : Dict = vae_state_dict["""encoder.conv_out.weight"""]
_A : Tuple = vae_state_dict["""encoder.conv_out.bias"""]
_A : List[Any] = vae_state_dict["""encoder.norm_out.weight"""]
_A : Tuple = vae_state_dict["""encoder.norm_out.bias"""]
_A : Optional[int] = vae_state_dict["""decoder.conv_in.weight"""]
_A : Tuple = vae_state_dict["""decoder.conv_in.bias"""]
_A : Union[str, Any] = vae_state_dict["""decoder.conv_out.weight"""]
_A : Tuple = vae_state_dict["""decoder.conv_out.bias"""]
_A : Tuple = vae_state_dict["""decoder.norm_out.weight"""]
_A : List[str] = vae_state_dict["""decoder.norm_out.bias"""]
_A : Optional[int] = vae_state_dict["""quant_conv.weight"""]
_A : List[Any] = vae_state_dict["""quant_conv.bias"""]
_A : Optional[Any] = vae_state_dict["""post_quant_conv.weight"""]
_A : int = vae_state_dict["""post_quant_conv.bias"""]
# Retrieves the keys for the encoder down blocks only
_A : Any = len({""".""".join(layer.split(""".""" )[:3] ) for layer in vae_state_dict if """encoder.down""" in layer} )
_A : Optional[Any] = {
layer_id: [key for key in vae_state_dict if f'''down.{layer_id}''' in key] for layer_id in range(snake_case_ )
}
# Retrieves the keys for the decoder up blocks only
_A : Any = len({""".""".join(layer.split(""".""" )[:3] ) for layer in vae_state_dict if """decoder.up""" in layer} )
_A : int = {
layer_id: [key for key in vae_state_dict if f'''up.{layer_id}''' in key] for layer_id in range(snake_case_ )
}
for i in range(snake_case_ ):
_A : str = [key for key in down_blocks[i] if f'''down.{i}''' in key and f'''down.{i}.downsample''' not in key]
if f'''encoder.down.{i}.downsample.conv.weight''' in vae_state_dict:
_A : Any = vae_state_dict.pop(
f'''encoder.down.{i}.downsample.conv.weight''' )
_A : List[str] = vae_state_dict.pop(
f'''encoder.down.{i}.downsample.conv.bias''' )
_A : Optional[Any] = renew_vae_resnet_paths(snake_case_ )
_A : str = {"""old""": f'''down.{i}.block''', """new""": f'''down_blocks.{i}.resnets'''}
assign_to_checkpoint(snake_case_,snake_case_,snake_case_,additional_replacements=[meta_path],config=snake_case_ )
_A : List[str] = [key for key in vae_state_dict if """encoder.mid.block""" in key]
_A : Union[str, Any] = 2
for i in range(1,num_mid_res_blocks + 1 ):
_A : str = [key for key in mid_resnets if f'''encoder.mid.block_{i}''' in key]
_A : str = renew_vae_resnet_paths(snake_case_ )
_A : str = {"""old""": f'''mid.block_{i}''', """new""": f'''mid_block.resnets.{i - 1}'''}
assign_to_checkpoint(snake_case_,snake_case_,snake_case_,additional_replacements=[meta_path],config=snake_case_ )
_A : str = [key for key in vae_state_dict if """encoder.mid.attn""" in key]
_A : int = renew_vae_attention_paths(snake_case_ )
_A : List[str] = {"""old""": """mid.attn_1""", """new""": """mid_block.attentions.0"""}
assign_to_checkpoint(snake_case_,snake_case_,snake_case_,additional_replacements=[meta_path],config=snake_case_ )
conv_attn_to_linear(snake_case_ )
for i in range(snake_case_ ):
_A : List[str] = num_up_blocks - 1 - i
_A : List[str] = [
key for key in up_blocks[block_id] if f'''up.{block_id}''' in key and f'''up.{block_id}.upsample''' not in key
]
if f'''decoder.up.{block_id}.upsample.conv.weight''' in vae_state_dict:
_A : Union[str, Any] = vae_state_dict[
f'''decoder.up.{block_id}.upsample.conv.weight'''
]
_A : List[str] = vae_state_dict[
f'''decoder.up.{block_id}.upsample.conv.bias'''
]
_A : str = renew_vae_resnet_paths(snake_case_ )
_A : int = {"""old""": f'''up.{block_id}.block''', """new""": f'''up_blocks.{i}.resnets'''}
assign_to_checkpoint(snake_case_,snake_case_,snake_case_,additional_replacements=[meta_path],config=snake_case_ )
_A : Tuple = [key for key in vae_state_dict if """decoder.mid.block""" in key]
_A : Tuple = 2
for i in range(1,num_mid_res_blocks + 1 ):
_A : int = [key for key in mid_resnets if f'''decoder.mid.block_{i}''' in key]
_A : Any = renew_vae_resnet_paths(snake_case_ )
_A : List[str] = {"""old""": f'''mid.block_{i}''', """new""": f'''mid_block.resnets.{i - 1}'''}
assign_to_checkpoint(snake_case_,snake_case_,snake_case_,additional_replacements=[meta_path],config=snake_case_ )
_A : Union[str, Any] = [key for key in vae_state_dict if """decoder.mid.attn""" in key]
_A : Any = renew_vae_attention_paths(snake_case_ )
_A : Optional[int] = {"""old""": """mid.attn_1""", """new""": """mid_block.attentions.0"""}
assign_to_checkpoint(snake_case_,snake_case_,snake_case_,additional_replacements=[meta_path],config=snake_case_ )
conv_attn_to_linear(snake_case_ )
return new_checkpoint
def lowerCAmelCase_ ( snake_case_,snake_case_,):
# Only support V1
_A : Tuple = requests.get(
""" https://raw.githubusercontent.com/CompVis/stable-diffusion/main/configs/stable-diffusion/v1-inference.yaml""" )
_A : Tuple = io.BytesIO(r.content )
_A : List[Any] = OmegaConf.load(snake_case_ )
_A : Optional[int] = 512
_A : Dict = """cuda""" if torch.cuda.is_available() else """cpu"""
if checkpoint_path.endswith("""safetensors""" ):
from safetensors import safe_open
_A : Tuple = {}
with safe_open(snake_case_,framework="""pt""",device="""cpu""" ) as f:
for key in f.keys():
_A : int = f.get_tensor(snake_case_ )
else:
_A : int = torch.load(snake_case_,map_location=snake_case_ )["""state_dict"""]
# Convert the VAE model.
_A : List[str] = create_vae_diffusers_config(snake_case_,image_size=snake_case_ )
_A : Any = custom_convert_ldm_vae_checkpoint(snake_case_,snake_case_ )
_A : List[Any] = AutoencoderKL(**snake_case_ )
vae.load_state_dict(snake_case_ )
vae.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
parser.add_argument("--vae_pt_path", default=None, type=str, required=True, help="Path to the VAE.pt to convert.")
parser.add_argument("--dump_path", default=None, type=str, required=True, help="Path to the VAE.pt to convert.")
_snake_case = parser.parse_args()
vae_pt_to_vae_diffuser(args.vae_pt_path, args.dump_path)
| 26 |
from __future__ import absolute_import, division, print_function, unicode_literals
from torch import nn
from torch.nn import CrossEntropyLoss, MSELoss
from transformers import RobertaConfig
from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
from transformers.models.roberta.modeling_roberta import (
ROBERTA_INPUTS_DOCSTRING,
ROBERTA_START_DOCSTRING,
RobertaEmbeddings,
)
from .modeling_highway_bert import BertPreTrainedModel, DeeBertModel, HighwayException, entropy
@add_start_docstrings(
"The RoBERTa Model transformer with early exiting (DeeRoBERTa). ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> Optional[int]:
super().__init__(_a )
_A : Union[str, Any] = RobertaEmbeddings(_a )
self.init_weights()
@add_start_docstrings(
"RoBERTa Model (with early exiting - DeeRoBERTa) with a classifier on top,\n also takes care of multi-layer training. ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> str:
super().__init__(_a )
_A : Any = config.num_labels
_A : Dict = config.num_hidden_layers
_A : List[str] = DeeRobertaModel(_a )
_A : int = nn.Dropout(config.hidden_dropout_prob )
_A : int = nn.Linear(config.hidden_size , self.config.num_labels )
@add_start_docstrings_to_model_forward(_a )
def a__ ( self , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=-1 , _a=False , ) -> Any:
_A : Optional[int] = self.num_layers
try:
_A : List[str] = self.roberta(
_a , attention_mask=_a , token_type_ids=_a , position_ids=_a , head_mask=_a , inputs_embeds=_a , )
_A : List[str] = outputs[1]
_A : List[str] = self.dropout(_a )
_A : Optional[Any] = self.classifier(_a )
_A : List[Any] = (logits,) + outputs[2:] # add hidden states and attention if they are here
except HighwayException as e:
_A : List[Any] = e.message
_A : Optional[int] = e.exit_layer
_A : Optional[int] = outputs[0]
if not self.training:
_A : int = entropy(_a )
_A : int = []
_A : int = []
if labels is not None:
if self.num_labels == 1:
# We are doing regression
_A : Union[str, Any] = MSELoss()
_A : Tuple = loss_fct(logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Dict = loss_fct(logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
# work with highway exits
_A : Optional[Any] = []
for highway_exit in outputs[-1]:
_A : Tuple = highway_exit[0]
if not self.training:
highway_logits_all.append(_a )
highway_entropy.append(highway_exit[2] )
if self.num_labels == 1:
# We are doing regression
_A : List[str] = MSELoss()
_A : Optional[int] = loss_fct(highway_logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Tuple = loss_fct(highway_logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
highway_losses.append(_a )
if train_highway:
_A : Dict = (sum(highway_losses[:-1] ),) + outputs
# exclude the final highway, of course
else:
_A : int = (loss,) + outputs
if not self.training:
_A : Optional[Any] = outputs + ((original_entropy, highway_entropy), exit_layer)
if output_layer >= 0:
_A : Union[str, Any] = (
(outputs[0],) + (highway_logits_all[output_layer],) + outputs[2:]
) # use the highway of the last layer
return outputs # (loss), logits, (hidden_states), (attentions), entropy
| 26 | 1 |
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_tf_available,
is_torch_available,
is_vision_available,
)
_snake_case = {
"configuration_blip": [
"BLIP_PRETRAINED_CONFIG_ARCHIVE_MAP",
"BlipConfig",
"BlipTextConfig",
"BlipVisionConfig",
],
"processing_blip": ["BlipProcessor"],
}
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = ["BlipImageProcessor"]
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"BLIP_PRETRAINED_MODEL_ARCHIVE_LIST",
"BlipModel",
"BlipPreTrainedModel",
"BlipForConditionalGeneration",
"BlipForQuestionAnswering",
"BlipVisionModel",
"BlipTextModel",
"BlipForImageTextRetrieval",
]
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"TF_BLIP_PRETRAINED_MODEL_ARCHIVE_LIST",
"TFBlipModel",
"TFBlipPreTrainedModel",
"TFBlipForConditionalGeneration",
"TFBlipForQuestionAnswering",
"TFBlipVisionModel",
"TFBlipTextModel",
"TFBlipForImageTextRetrieval",
]
if TYPE_CHECKING:
from .configuration_blip import BLIP_PRETRAINED_CONFIG_ARCHIVE_MAP, BlipConfig, BlipTextConfig, BlipVisionConfig
from .processing_blip import BlipProcessor
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .image_processing_blip import BlipImageProcessor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_blip import (
BLIP_PRETRAINED_MODEL_ARCHIVE_LIST,
BlipForConditionalGeneration,
BlipForImageTextRetrieval,
BlipForQuestionAnswering,
BlipModel,
BlipPreTrainedModel,
BlipTextModel,
BlipVisionModel,
)
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_tf_blip import (
TF_BLIP_PRETRAINED_MODEL_ARCHIVE_LIST,
TFBlipForConditionalGeneration,
TFBlipForImageTextRetrieval,
TFBlipForQuestionAnswering,
TFBlipModel,
TFBlipPreTrainedModel,
TFBlipTextModel,
TFBlipVisionModel,
)
else:
import sys
_snake_case = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
| 26 |
from collections import OrderedDict
from typing import Mapping
from ...configuration_utils import PretrainedConfig
from ...onnx import OnnxConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"facebook/xmod-base": "https://huggingface.co/facebook/xmod-base/resolve/main/config.json",
"facebook/xmod-large-prenorm": "https://huggingface.co/facebook/xmod-large-prenorm/resolve/main/config.json",
"facebook/xmod-base-13-125k": "https://huggingface.co/facebook/xmod-base-13-125k/resolve/main/config.json",
"facebook/xmod-base-30-125k": "https://huggingface.co/facebook/xmod-base-30-125k/resolve/main/config.json",
"facebook/xmod-base-30-195k": "https://huggingface.co/facebook/xmod-base-30-195k/resolve/main/config.json",
"facebook/xmod-base-60-125k": "https://huggingface.co/facebook/xmod-base-60-125k/resolve/main/config.json",
"facebook/xmod-base-60-265k": "https://huggingface.co/facebook/xmod-base-60-265k/resolve/main/config.json",
"facebook/xmod-base-75-125k": "https://huggingface.co/facebook/xmod-base-75-125k/resolve/main/config.json",
"facebook/xmod-base-75-269k": "https://huggingface.co/facebook/xmod-base-75-269k/resolve/main/config.json",
}
class lowercase ( UpperCamelCase__ ):
_a = "xmod"
def __init__( self , _a=3_0522 , _a=768 , _a=12 , _a=12 , _a=3072 , _a="gelu" , _a=0.1 , _a=0.1 , _a=512 , _a=2 , _a=0.02 , _a=1e-12 , _a=1 , _a=0 , _a=2 , _a="absolute" , _a=True , _a=None , _a=False , _a=2 , _a=False , _a=True , _a=True , _a=("en_XX",) , _a=None , **_a , ) -> str:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Tuple = vocab_size
_A : Union[str, Any] = hidden_size
_A : Dict = num_hidden_layers
_A : Dict = num_attention_heads
_A : List[Any] = hidden_act
_A : Optional[Any] = intermediate_size
_A : Any = hidden_dropout_prob
_A : str = attention_probs_dropout_prob
_A : Dict = max_position_embeddings
_A : Any = type_vocab_size
_A : List[Any] = initializer_range
_A : int = layer_norm_eps
_A : int = position_embedding_type
_A : Any = use_cache
_A : int = classifier_dropout
_A : int = pre_norm
_A : Optional[Any] = adapter_reduction_factor
_A : List[Any] = adapter_layer_norm
_A : Optional[int] = adapter_reuse_layer_norm
_A : Any = ln_before_adapter
_A : Union[str, Any] = list(_a )
_A : List[Any] = default_language
class lowercase ( UpperCamelCase__ ):
@property
def a__ ( self ) -> Mapping[str, Mapping[int, str]]:
if self.task == "multiple-choice":
_A : Dict = {0: """batch""", 1: """choice""", 2: """sequence"""}
else:
_A : Dict = {0: """batch""", 1: """sequence"""}
return OrderedDict(
[
("""input_ids""", dynamic_axis),
("""attention_mask""", dynamic_axis),
] )
| 26 | 1 |
import shutil
import tempfile
import unittest
import numpy as np
import pytest
from transformers.testing_utils import require_vision
from transformers.utils import is_vision_available
if is_vision_available():
from PIL import Image
from transformers import AutoProcessor, BlipaProcessor, BlipImageProcessor, GPTaTokenizer, PreTrainedTokenizerFast
@require_vision
class lowercase ( unittest.TestCase ):
def a__ ( self ) -> Any:
_A : str = tempfile.mkdtemp()
_A : str = BlipImageProcessor()
_A : List[Any] = GPTaTokenizer.from_pretrained("""hf-internal-testing/tiny-random-GPT2Model""" )
_A : Union[str, Any] = BlipaProcessor(_a , _a )
processor.save_pretrained(self.tmpdirname )
def a__ ( self , **_a ) -> List[Any]:
return AutoProcessor.from_pretrained(self.tmpdirname , **_a ).tokenizer
def a__ ( self , **_a ) -> Optional[int]:
return AutoProcessor.from_pretrained(self.tmpdirname , **_a ).image_processor
def a__ ( self ) -> Any:
shutil.rmtree(self.tmpdirname )
def a__ ( self ) -> Any:
_A : List[str] = [np.random.randint(255 , size=(3, 30, 400) , dtype=np.uinta )]
_A : List[str] = [Image.fromarray(np.moveaxis(_a , 0 , -1 ) ) for x in image_inputs]
return image_inputs
def a__ ( self ) -> List[Any]:
_A : List[Any] = BlipaProcessor(tokenizer=self.get_tokenizer() , image_processor=self.get_image_processor() )
processor.save_pretrained(self.tmpdirname )
_A : Dict = self.get_tokenizer(bos_token="""(BOS)""" , eos_token="""(EOS)""" )
_A : Optional[int] = self.get_image_processor(do_normalize=_a , padding_value=1.0 )
_A : List[Any] = BlipaProcessor.from_pretrained(
self.tmpdirname , bos_token="""(BOS)""" , eos_token="""(EOS)""" , do_normalize=_a , padding_value=1.0 )
self.assertEqual(processor.tokenizer.get_vocab() , tokenizer_add_kwargs.get_vocab() )
self.assertIsInstance(processor.tokenizer , _a )
self.assertEqual(processor.image_processor.to_json_string() , image_processor_add_kwargs.to_json_string() )
self.assertIsInstance(processor.image_processor , _a )
def a__ ( self ) -> str:
_A : Any = self.get_image_processor()
_A : Optional[Any] = self.get_tokenizer()
_A : List[Any] = BlipaProcessor(tokenizer=_a , image_processor=_a )
_A : List[str] = self.prepare_image_inputs()
_A : Optional[Any] = image_processor(_a , return_tensors="""np""" )
_A : Optional[int] = processor(images=_a , return_tensors="""np""" )
for key in input_feat_extract.keys():
self.assertAlmostEqual(input_feat_extract[key].sum() , input_processor[key].sum() , delta=1e-2 )
def a__ ( self ) -> List[Any]:
_A : Tuple = self.get_image_processor()
_A : List[str] = self.get_tokenizer()
_A : str = BlipaProcessor(tokenizer=_a , image_processor=_a )
_A : Optional[Any] = """lower newer"""
_A : Tuple = processor(text=_a )
_A : Optional[int] = tokenizer(_a , return_token_type_ids=_a )
for key in encoded_tok.keys():
self.assertListEqual(encoded_tok[key] , encoded_processor[key] )
def a__ ( self ) -> Optional[Any]:
_A : str = self.get_image_processor()
_A : List[str] = self.get_tokenizer()
_A : Dict = BlipaProcessor(tokenizer=_a , image_processor=_a )
_A : int = """lower newer"""
_A : List[str] = self.prepare_image_inputs()
_A : Any = processor(text=_a , images=_a )
self.assertListEqual(list(inputs.keys() ) , ["""pixel_values""", """input_ids""", """attention_mask"""] )
# test if it raises when no input is passed
with pytest.raises(_a ):
processor()
def a__ ( self ) -> Optional[Any]:
_A : List[Any] = self.get_image_processor()
_A : Optional[int] = self.get_tokenizer()
_A : str = BlipaProcessor(tokenizer=_a , image_processor=_a )
_A : Union[str, Any] = [[1, 4, 5, 8, 1, 0, 8], [3, 4, 3, 1, 1, 8, 9]]
_A : Tuple = processor.batch_decode(_a )
_A : Optional[Any] = tokenizer.batch_decode(_a )
self.assertListEqual(_a , _a )
def a__ ( self ) -> Any:
_A : List[str] = self.get_image_processor()
_A : Any = self.get_tokenizer()
_A : Optional[Any] = BlipaProcessor(tokenizer=_a , image_processor=_a )
_A : Tuple = """lower newer"""
_A : str = self.prepare_image_inputs()
_A : Any = processor(text=_a , images=_a )
# For now the processor supports only ['pixel_values', 'input_ids', 'attention_mask']
self.assertListEqual(list(inputs.keys() ) , ["""pixel_values""", """input_ids""", """attention_mask"""] )
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
if n == 0:
return 0
_A : Tuple = float("""-inf""" )
for i in range(1,n + 1 ):
_A : str = max(
snake_case_,prices[i - 1] + naive_cut_rod_recursive(n - i,snake_case_ ) )
return max_revue
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
_A : Dict = [float("""-inf""" ) for _ in range(n + 1 )]
return _top_down_cut_rod_recursive(snake_case_,snake_case_,snake_case_ )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
if max_rev[n] >= 0:
return max_rev[n]
elif n == 0:
return 0
else:
_A : List[str] = float("""-inf""" )
for i in range(1,n + 1 ):
_A : Optional[Any] = max(
snake_case_,prices[i - 1] + _top_down_cut_rod_recursive(n - i,snake_case_,snake_case_ ),)
_A : Tuple = max_revenue
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
# length(max_rev) = n + 1, to accommodate for the revenue obtainable from a rod of
# length 0.
_A : List[Any] = [float("""-inf""" ) for _ in range(n + 1 )]
_A : Any = 0
for i in range(1,n + 1 ):
_A : Optional[Any] = max_rev[i]
for j in range(1,i + 1 ):
_A : int = max(snake_case_,prices[j - 1] + max_rev[i - j] )
_A : int = max_revenue_i
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if n < 0:
_A : Optional[Any] = f'''n must be greater than or equal to 0. Got n = {n}'''
raise ValueError(snake_case_ )
if n > len(snake_case_ ):
_A : Any = (
"""Each integral piece of rod must have a corresponding price. """
f'''Got n = {n} but length of prices = {len(snake_case_ )}'''
)
raise ValueError(snake_case_ )
def lowerCAmelCase_ ( ):
_A : Tuple = [6, 10, 12, 15, 20, 23]
_A : List[Any] = len(snake_case_ )
# the best revenue comes from cutting the rod into 6 pieces, each
# of length 1 resulting in a revenue of 6 * 6 = 36.
_A : Any = 36
_A : List[Any] = top_down_cut_rod(snake_case_,snake_case_ )
_A : List[Any] = bottom_up_cut_rod(snake_case_,snake_case_ )
_A : Dict = naive_cut_rod_recursive(snake_case_,snake_case_ )
assert expected_max_revenue == max_rev_top_down
assert max_rev_top_down == max_rev_bottom_up
assert max_rev_bottom_up == max_rev_naive
if __name__ == "__main__":
main()
| 26 | 1 |
from __future__ import absolute_import, division, print_function, unicode_literals
from torch import nn
from torch.nn import CrossEntropyLoss, MSELoss
from transformers import RobertaConfig
from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
from transformers.models.roberta.modeling_roberta import (
ROBERTA_INPUTS_DOCSTRING,
ROBERTA_START_DOCSTRING,
RobertaEmbeddings,
)
from .modeling_highway_bert import BertPreTrainedModel, DeeBertModel, HighwayException, entropy
@add_start_docstrings(
"The RoBERTa Model transformer with early exiting (DeeRoBERTa). ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> Optional[int]:
super().__init__(_a )
_A : Union[str, Any] = RobertaEmbeddings(_a )
self.init_weights()
@add_start_docstrings(
"RoBERTa Model (with early exiting - DeeRoBERTa) with a classifier on top,\n also takes care of multi-layer training. ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> str:
super().__init__(_a )
_A : Any = config.num_labels
_A : Dict = config.num_hidden_layers
_A : List[str] = DeeRobertaModel(_a )
_A : int = nn.Dropout(config.hidden_dropout_prob )
_A : int = nn.Linear(config.hidden_size , self.config.num_labels )
@add_start_docstrings_to_model_forward(_a )
def a__ ( self , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=-1 , _a=False , ) -> Any:
_A : Optional[int] = self.num_layers
try:
_A : List[str] = self.roberta(
_a , attention_mask=_a , token_type_ids=_a , position_ids=_a , head_mask=_a , inputs_embeds=_a , )
_A : List[str] = outputs[1]
_A : List[str] = self.dropout(_a )
_A : Optional[Any] = self.classifier(_a )
_A : List[Any] = (logits,) + outputs[2:] # add hidden states and attention if they are here
except HighwayException as e:
_A : List[Any] = e.message
_A : Optional[int] = e.exit_layer
_A : Optional[int] = outputs[0]
if not self.training:
_A : int = entropy(_a )
_A : int = []
_A : int = []
if labels is not None:
if self.num_labels == 1:
# We are doing regression
_A : Union[str, Any] = MSELoss()
_A : Tuple = loss_fct(logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Dict = loss_fct(logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
# work with highway exits
_A : Optional[Any] = []
for highway_exit in outputs[-1]:
_A : Tuple = highway_exit[0]
if not self.training:
highway_logits_all.append(_a )
highway_entropy.append(highway_exit[2] )
if self.num_labels == 1:
# We are doing regression
_A : List[str] = MSELoss()
_A : Optional[int] = loss_fct(highway_logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Tuple = loss_fct(highway_logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
highway_losses.append(_a )
if train_highway:
_A : Dict = (sum(highway_losses[:-1] ),) + outputs
# exclude the final highway, of course
else:
_A : int = (loss,) + outputs
if not self.training:
_A : Optional[Any] = outputs + ((original_entropy, highway_entropy), exit_layer)
if output_layer >= 0:
_A : Union[str, Any] = (
(outputs[0],) + (highway_logits_all[output_layer],) + outputs[2:]
) # use the highway of the last layer
return outputs # (loss), logits, (hidden_states), (attentions), entropy
| 26 |
import requests
from bsa import BeautifulSoup
def lowerCAmelCase_ ( snake_case_ = "AAPL" ):
_A : str = f'''https://in.finance.yahoo.com/quote/{symbol}?s={symbol}'''
_A : List[Any] = BeautifulSoup(requests.get(snake_case_ ).text,"""html.parser""" )
_A : Union[str, Any] = """My(6px) Pos(r) smartphone_Mt(6px)"""
return soup.find("""div""",class_=class_ ).find("""span""" ).text
if __name__ == "__main__":
for symbol in "AAPL AMZN IBM GOOG MSFT ORCL".split():
print(f"""Current {symbol:<4} stock price is {stock_price(symbol):>8}""")
| 26 | 1 |
from typing import Any
import numpy as np
def lowerCAmelCase_ ( snake_case_ ):
return np.array_equal(snake_case_,matrix.conjugate().T )
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Union[str, Any] = v.conjugate().T
_A : Any = v_star.dot(snake_case_ )
assert isinstance(snake_case_,np.ndarray )
return (v_star_dot.dot(snake_case_ )) / (v_star.dot(snake_case_ ))
def lowerCAmelCase_ ( ):
_A : str = np.array([[2, 2 + 1J, 4], [2 - 1J, 3, 1J], [4, -1J, 1]] )
_A : str = np.array([[1], [2], [3]] )
assert is_hermitian(snake_case_ ), f'''{a} is not hermitian.'''
print(rayleigh_quotient(snake_case_,snake_case_ ) )
_A : List[str] = np.array([[1, 2, 4], [2, 3, -1], [4, -1, 1]] )
assert is_hermitian(snake_case_ ), f'''{a} is not hermitian.'''
assert rayleigh_quotient(snake_case_,snake_case_ ) == float(3 )
if __name__ == "__main__":
import doctest
doctest.testmod()
tests()
| 26 |
import unittest
from transformers import (
MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
TextaTextGenerationPipeline,
pipeline,
)
from transformers.testing_utils import is_pipeline_test, require_tf, require_torch
from transformers.utils import is_torch_available
from .test_pipelines_common import ANY
if is_torch_available():
import torch
@is_pipeline_test
class lowercase ( unittest.TestCase ):
_a = MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING
_a = TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING
def a__ ( self , _a , _a , _a ) -> int:
_A : str = TextaTextGenerationPipeline(model=_a , tokenizer=_a )
return generator, ["Something to write", "Something else"]
def a__ ( self , _a , _a ) -> Dict:
_A : Any = generator("""Something there""" )
self.assertEqual(_a , [{"""generated_text""": ANY(_a )}] )
# These are encoder decoder, they don't just append to incoming string
self.assertFalse(outputs[0]["""generated_text"""].startswith("""Something there""" ) )
_A : List[Any] = generator(["""This is great !""", """Something else"""] , num_return_sequences=2 , do_sample=_a )
self.assertEqual(
_a , [
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
] , )
_A : Optional[int] = generator(
["""This is great !""", """Something else"""] , num_return_sequences=2 , batch_size=2 , do_sample=_a )
self.assertEqual(
_a , [
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
] , )
with self.assertRaises(_a ):
generator(4 )
@require_torch
def a__ ( self ) -> List[str]:
_A : Any = pipeline("""text2text-generation""" , model="""patrickvonplaten/t5-tiny-random""" , framework="""pt""" )
# do_sample=False necessary for reproducibility
_A : Dict = generator("""Something there""" , do_sample=_a )
self.assertEqual(_a , [{"""generated_text""": """"""}] )
_A : Any = 3
_A : Any = generator(
"""Something there""" , num_return_sequences=_a , num_beams=_a , )
_A : Optional[int] = [
{"""generated_text""": """Beide Beide Beide Beide Beide Beide Beide Beide Beide"""},
{"""generated_text""": """Beide Beide Beide Beide Beide Beide Beide Beide"""},
{"""generated_text""": """"""},
]
self.assertEqual(_a , _a )
_A : Dict = generator("""This is a test""" , do_sample=_a , num_return_sequences=2 , return_tensors=_a )
self.assertEqual(
_a , [
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
] , )
_A : Dict = generator.model.config.eos_token_id
_A : List[str] = """<pad>"""
_A : Dict = generator(
["""This is a test""", """This is a second test"""] , do_sample=_a , num_return_sequences=2 , batch_size=2 , return_tensors=_a , )
self.assertEqual(
_a , [
[
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
],
[
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
],
] , )
@require_tf
def a__ ( self ) -> int:
_A : Optional[Any] = pipeline("""text2text-generation""" , model="""patrickvonplaten/t5-tiny-random""" , framework="""tf""" )
# do_sample=False necessary for reproducibility
_A : str = generator("""Something there""" , do_sample=_a )
self.assertEqual(_a , [{"""generated_text""": """"""}] )
| 26 | 1 |
# Copyright 2021 The HuggingFace Team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import argparse
import os
from accelerate.test_utils import execute_subprocess_async
def lowerCAmelCase_ ( snake_case_=None ):
if subparsers is not None:
_A : List[str] = subparsers.add_parser("""test""" )
else:
_A : str = argparse.ArgumentParser("""Accelerate test command""" )
parser.add_argument(
"""--config_file""",default=snake_case_,help=(
"""The path to use to store the config file. Will default to a file named default_config.yaml in the cache """
"""location, which is the content of the environment `HF_HOME` suffixed with 'accelerate', or if you don't have """
"""such an environment variable, your cache directory ('~/.cache' or the content of `XDG_CACHE_HOME`) suffixed """
"""with 'huggingface'."""
),)
if subparsers is not None:
parser.set_defaults(func=snake_case_ )
return parser
def lowerCAmelCase_ ( snake_case_ ):
_A : str = os.path.sep.join(__file__.split(os.path.sep )[:-2] + ["""test_utils""", """scripts""", """test_script.py"""] )
if args.config_file is None:
_A : Tuple = script_name
else:
_A : Tuple = f'''--config_file={args.config_file} {script_name}'''
_A : int = ["""accelerate-launch"""] + test_args.split()
_A : str = execute_subprocess_async(snake_case_,env=os.environ.copy() )
if result.returncode == 0:
print("""Test is a success! You are ready for your distributed training!""" )
def lowerCAmelCase_ ( ):
_A : List[Any] = test_command_parser()
_A : Any = parser.parse_args()
test_command(snake_case_ )
if __name__ == "__main__":
main()
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
while b:
_A , _A : List[str] = b, a % b
return a
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
return a if b == 0 else euclidean_gcd_recursive(snake_case_,a % b )
def lowerCAmelCase_ ( ):
print(f'''euclidean_gcd(3, 5) = {euclidean_gcd(3,5 )}''' )
print(f'''euclidean_gcd(5, 3) = {euclidean_gcd(5,3 )}''' )
print(f'''euclidean_gcd(1, 3) = {euclidean_gcd(1,3 )}''' )
print(f'''euclidean_gcd(3, 6) = {euclidean_gcd(3,6 )}''' )
print(f'''euclidean_gcd(6, 3) = {euclidean_gcd(6,3 )}''' )
print(f'''euclidean_gcd_recursive(3, 5) = {euclidean_gcd_recursive(3,5 )}''' )
print(f'''euclidean_gcd_recursive(5, 3) = {euclidean_gcd_recursive(5,3 )}''' )
print(f'''euclidean_gcd_recursive(1, 3) = {euclidean_gcd_recursive(1,3 )}''' )
print(f'''euclidean_gcd_recursive(3, 6) = {euclidean_gcd_recursive(3,6 )}''' )
print(f'''euclidean_gcd_recursive(6, 3) = {euclidean_gcd_recursive(6,3 )}''' )
if __name__ == "__main__":
main()
| 26 | 1 |
_snake_case = "0.21.0"
from .accelerator import Accelerator
from .big_modeling import (
cpu_offload,
cpu_offload_with_hook,
disk_offload,
dispatch_model,
init_empty_weights,
init_on_device,
load_checkpoint_and_dispatch,
)
from .data_loader import skip_first_batches
from .launchers import debug_launcher, notebook_launcher
from .state import PartialState
from .utils import (
DeepSpeedPlugin,
DistributedDataParallelKwargs,
DistributedType,
FullyShardedDataParallelPlugin,
GradScalerKwargs,
InitProcessGroupKwargs,
find_executable_batch_size,
infer_auto_device_map,
is_rich_available,
load_checkpoint_in_model,
synchronize_rng_states,
)
if is_rich_available():
from .utils import rich
| 26 |
def lowerCAmelCase_ ( snake_case_ ):
if number < 0:
raise ValueError("""number must not be negative""" )
return number & (number - 1) == 0
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
import argparse
import json
from pathlib import Path
import requests
import torch
from huggingface_hub import hf_hub_download
from PIL import Image
from transformers import YolosConfig, YolosForObjectDetection, YolosImageProcessor
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
def lowerCAmelCase_ ( snake_case_ ):
_A : Union[str, Any] = YolosConfig()
# size of the architecture
if "yolos_ti" in yolos_name:
_A : Union[str, Any] = 192
_A : str = 768
_A : List[Any] = 12
_A : str = 3
_A : int = [800, 1333]
_A : Union[str, Any] = False
elif yolos_name == "yolos_s_dWr":
_A : List[str] = 330
_A : Optional[Any] = 14
_A : List[Any] = 6
_A : Optional[int] = 1320
elif "yolos_s" in yolos_name:
_A : Tuple = 384
_A : Any = 1536
_A : Any = 12
_A : List[Any] = 6
elif "yolos_b" in yolos_name:
_A : int = [800, 1344]
_A : Tuple = 91
_A : str = """huggingface/label-files"""
_A : Tuple = """coco-detection-id2label.json"""
_A : Any = json.load(open(hf_hub_download(snake_case_,snake_case_,repo_type="""dataset""" ),"""r""" ) )
_A : Dict = {int(snake_case_ ): v for k, v in idalabel.items()}
_A : Tuple = idalabel
_A : Any = {v: k for k, v in idalabel.items()}
return config
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ = False ):
for i in range(config.num_hidden_layers ):
# read in weights + bias of input projection layer (in timm, this is a single matrix + bias)
_A : List[str] = state_dict.pop(f'''blocks.{i}.attn.qkv.weight''' )
_A : int = state_dict.pop(f'''blocks.{i}.attn.qkv.bias''' )
# next, add query, keys and values (in that order) to the state dict
_A : Union[str, Any] = in_proj_weight[: config.hidden_size, :]
_A : Union[str, Any] = in_proj_bias[: config.hidden_size]
_A : Tuple = in_proj_weight[
config.hidden_size : config.hidden_size * 2, :
]
_A : List[Any] = in_proj_bias[
config.hidden_size : config.hidden_size * 2
]
_A : Any = in_proj_weight[-config.hidden_size :, :]
_A : Optional[int] = in_proj_bias[-config.hidden_size :]
def lowerCAmelCase_ ( snake_case_ ):
if "backbone" in name:
_A : str = name.replace("""backbone""","""vit""" )
if "cls_token" in name:
_A : Any = name.replace("""cls_token""","""embeddings.cls_token""" )
if "det_token" in name:
_A : int = name.replace("""det_token""","""embeddings.detection_tokens""" )
if "mid_pos_embed" in name:
_A : int = name.replace("""mid_pos_embed""","""encoder.mid_position_embeddings""" )
if "pos_embed" in name:
_A : Optional[int] = name.replace("""pos_embed""","""embeddings.position_embeddings""" )
if "patch_embed.proj" in name:
_A : str = name.replace("""patch_embed.proj""","""embeddings.patch_embeddings.projection""" )
if "blocks" in name:
_A : Union[str, Any] = name.replace("""blocks""","""encoder.layer""" )
if "attn.proj" in name:
_A : Tuple = name.replace("""attn.proj""","""attention.output.dense""" )
if "attn" in name:
_A : str = name.replace("""attn""","""attention.self""" )
if "norm1" in name:
_A : Tuple = name.replace("""norm1""","""layernorm_before""" )
if "norm2" in name:
_A : Optional[Any] = name.replace("""norm2""","""layernorm_after""" )
if "mlp.fc1" in name:
_A : Optional[int] = name.replace("""mlp.fc1""","""intermediate.dense""" )
if "mlp.fc2" in name:
_A : Dict = name.replace("""mlp.fc2""","""output.dense""" )
if "class_embed" in name:
_A : Union[str, Any] = name.replace("""class_embed""","""class_labels_classifier""" )
if "bbox_embed" in name:
_A : Any = name.replace("""bbox_embed""","""bbox_predictor""" )
if "vit.norm" in name:
_A : Union[str, Any] = name.replace("""vit.norm""","""vit.layernorm""" )
return name
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
for key in orig_state_dict.copy().keys():
_A : Tuple = orig_state_dict.pop(snake_case_ )
if "qkv" in key:
_A : str = key.split(""".""" )
_A : Tuple = int(key_split[2] )
_A : Tuple = model.vit.encoder.layer[layer_num].attention.attention.all_head_size
if "weight" in key:
_A : Optional[Any] = val[:dim, :]
_A : str = val[
dim : dim * 2, :
]
_A : List[str] = val[-dim:, :]
else:
_A : Optional[int] = val[:dim]
_A : Union[str, Any] = val[dim : dim * 2]
_A : Tuple = val[-dim:]
else:
_A : List[Any] = val
return orig_state_dict
def lowerCAmelCase_ ( ):
_A : Any = """http://images.cocodataset.org/val2017/000000039769.jpg"""
_A : Optional[Any] = Image.open(requests.get(snake_case_,stream=snake_case_ ).raw )
return im
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ = False ):
_A : Optional[int] = get_yolos_config(snake_case_ )
# load original state_dict
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )["""model"""]
# load 🤗 model
_A : Tuple = YolosForObjectDetection(snake_case_ )
model.eval()
_A : Optional[int] = convert_state_dict(snake_case_,snake_case_ )
model.load_state_dict(snake_case_ )
# Check outputs on an image, prepared by YolosImageProcessor
_A : List[str] = 800 if yolos_name != """yolos_ti""" else 512
_A : Optional[Any] = YolosImageProcessor(format="""coco_detection""",size=snake_case_ )
_A : Dict = image_processor(images=prepare_img(),return_tensors="""pt""" )
_A : List[Any] = model(**snake_case_ )
_A , _A : Optional[int] = outputs.logits, outputs.pred_boxes
_A , _A : int = None, None
if yolos_name == "yolos_ti":
_A : Any = torch.tensor(
[[-39.50_22, -11.98_20, -17.68_88], [-29.95_74, -9.97_69, -17.76_91], [-42.32_81, -20.72_00, -30.62_94]] )
_A : Optional[Any] = torch.tensor(
[[0.40_21, 0.08_36, 0.79_79], [0.01_84, 0.26_09, 0.03_64], [0.17_81, 0.20_04, 0.20_95]] )
elif yolos_name == "yolos_s_200_pre":
_A : Union[str, Any] = torch.tensor(
[[-24.02_48, -10.30_24, -14.82_90], [-42.03_92, -16.82_00, -27.43_34], [-27.27_43, -11.81_54, -18.71_48]] )
_A : str = torch.tensor(
[[0.25_59, 0.54_55, 0.47_06], [0.29_89, 0.72_79, 0.18_75], [0.77_32, 0.40_17, 0.44_62]] )
elif yolos_name == "yolos_s_300_pre":
_A : str = torch.tensor(
[[-36.22_20, -14.43_85, -23.54_57], [-35.69_70, -14.75_83, -21.39_35], [-31.59_39, -13.60_42, -16.80_49]] )
_A : Optional[int] = torch.tensor(
[[0.76_14, 0.23_16, 0.47_28], [0.71_68, 0.44_95, 0.38_55], [0.49_96, 0.14_66, 0.99_96]] )
elif yolos_name == "yolos_s_dWr":
_A : str = torch.tensor(
[[-42.86_68, -24.10_49, -41.16_90], [-34.74_56, -14.12_74, -24.91_94], [-33.78_98, -12.19_46, -25.64_95]] )
_A : Any = torch.tensor(
[[0.55_87, 0.27_73, 0.06_05], [0.50_04, 0.30_14, 0.99_94], [0.49_99, 0.15_48, 0.99_94]] )
elif yolos_name == "yolos_base":
_A : Any = torch.tensor(
[[-40.60_64, -24.30_84, -32.64_47], [-55.19_90, -30.77_19, -35.58_77], [-51.43_11, -33.35_07, -35.64_62]] )
_A : List[str] = torch.tensor(
[[0.55_55, 0.27_94, 0.06_55], [0.90_49, 0.26_64, 0.18_94], [0.91_83, 0.19_84, 0.16_35]] )
else:
raise ValueError(f'''Unknown yolos_name: {yolos_name}''' )
assert torch.allclose(logits[0, :3, :3],snake_case_,atol=1e-4 )
assert torch.allclose(pred_boxes[0, :3, :3],snake_case_,atol=1e-4 )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ )
print(f'''Saving model {yolos_name} to {pytorch_dump_folder_path}''' )
model.save_pretrained(snake_case_ )
print(f'''Saving image processor to {pytorch_dump_folder_path}''' )
image_processor.save_pretrained(snake_case_ )
if push_to_hub:
_A : Optional[int] = {
"""yolos_ti""": """yolos-tiny""",
"""yolos_s_200_pre""": """yolos-small""",
"""yolos_s_300_pre""": """yolos-small-300""",
"""yolos_s_dWr""": """yolos-small-dwr""",
"""yolos_base""": """yolos-base""",
}
print("""Pushing to the hub...""" )
_A : int = model_mapping[yolos_name]
image_processor.push_to_hub(snake_case_,organization="""hustvl""" )
model.push_to_hub(snake_case_,organization="""hustvl""" )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
"--yolos_name",
default="yolos_s_200_pre",
type=str,
help=(
"Name of the YOLOS model you'd like to convert. Should be one of 'yolos_ti', 'yolos_s_200_pre',"
" 'yolos_s_300_pre', 'yolos_s_dWr', 'yolos_base'."
),
)
parser.add_argument(
"--checkpoint_path", default=None, type=str, help="Path to the original state dict (.pth file)."
)
parser.add_argument(
"--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model directory."
)
parser.add_argument(
"--push_to_hub", action="store_true", help="Whether or not to push the converted model to the 🤗 hub."
)
_snake_case = parser.parse_args()
convert_yolos_checkpoint(args.yolos_name, args.checkpoint_path, args.pytorch_dump_folder_path, args.push_to_hub)
| 26 |
import argparse
import torch
from transformers import BlenderbotConfig, BlenderbotForConditionalGeneration
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
_snake_case = [
["attention", "attn"],
["encoder_attention", "encoder_attn"],
["q_lin", "q_proj"],
["k_lin", "k_proj"],
["v_lin", "v_proj"],
["out_lin", "out_proj"],
["norm_embeddings", "layernorm_embedding"],
["position_embeddings", "embed_positions"],
["embeddings", "embed_tokens"],
["ffn.lin", "fc"],
]
def lowerCAmelCase_ ( snake_case_ ):
if k == "embeddings.weight":
return "shared.weight"
for parlai_name, hf_name in PATTERNS:
_A : str = k.replace(snake_case_,snake_case_ )
if k.startswith("""encoder""" ):
_A : Optional[Any] = k.replace(""".attn""",""".self_attn""" )
_A : Dict = k.replace("""norm1""","""self_attn_layer_norm""" )
_A : Optional[Any] = k.replace("""norm2""","""final_layer_norm""" )
elif k.startswith("""decoder""" ):
_A : str = k.replace("""norm1""","""self_attn_layer_norm""" )
_A : Any = k.replace("""norm2""","""encoder_attn_layer_norm""" )
_A : Optional[int] = k.replace("""norm3""","""final_layer_norm""" )
return k
def lowerCAmelCase_ ( snake_case_ ):
_A : List[Any] = [
"""model.encoder.layernorm_embedding.weight""",
"""model.encoder.layernorm_embedding.bias""",
"""model.decoder.layernorm_embedding.weight""",
"""model.decoder.layernorm_embedding.bias""",
]
for k in keys:
_A : str = sd.pop(snake_case_ )
_A : Optional[int] = k.replace("""layernorm_embedding""","""layer_norm""" )
assert new_k not in sd
_A : Optional[int] = v
_snake_case = ["START"]
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
_A : List[Any] = model["""model"""]
_A : Optional[Any] = BlenderbotConfig.from_json_file(snake_case_ )
_A : List[str] = BlenderbotForConditionalGeneration(snake_case_ )
_A : Tuple = m.model.state_dict().keys()
_A : Any = []
_A : Dict = {}
for k, v in sd.items():
if k in IGNORE_KEYS:
continue
_A : Optional[int] = rename_state_dict_key(snake_case_ )
if new_k not in valid_keys:
failures.append([k, new_k] )
else:
_A : Dict = v
if cfg.normalize_before: # Blenderbot-3B checkpoints. Rename layernorm_embedding -> layer_norm
rename_layernorm_keys(snake_case_ )
m.model.load_state_dict(snake_case_,strict=snake_case_ )
m.half()
m.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument("--src_path", type=str, help="like blenderbot-model.bin")
parser.add_argument("--save_dir", default="hf_blenderbot", type=str, help="Where to save converted model.")
parser.add_argument(
"--hf_config_json", default="blenderbot-3b-config.json", type=str, help="Path to config to use"
)
_snake_case = parser.parse_args()
convert_parlai_checkpoint(args.src_path, args.save_dir, args.hf_config_json)
| 26 | 1 |
from __future__ import annotations
import inspect
import unittest
from typing import List, Tuple
from transformers import RegNetConfig
from transformers.testing_utils import require_tf, require_vision, slow
from transformers.utils import cached_property, is_tf_available, is_vision_available
from ...test_configuration_common import ConfigTester
from ...test_modeling_tf_common import TFModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_tf_available():
import tensorflow as tf
from transformers import TF_REGNET_PRETRAINED_MODEL_ARCHIVE_LIST, TFRegNetForImageClassification, TFRegNetModel
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class lowercase :
def __init__( self , _a , _a=3 , _a=32 , _a=3 , _a=10 , _a=[10, 20, 30, 40] , _a=[1, 1, 2, 1] , _a=True , _a=True , _a="relu" , _a=3 , _a=None , ) -> Any:
_A : Tuple = parent
_A : int = batch_size
_A : int = image_size
_A : List[str] = num_channels
_A : Optional[int] = embeddings_size
_A : int = hidden_sizes
_A : Any = depths
_A : Dict = is_training
_A : Union[str, Any] = use_labels
_A : List[str] = hidden_act
_A : Any = num_labels
_A : int = scope
_A : Any = len(_a )
def a__ ( self ) -> Tuple:
_A : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
_A : Optional[Any] = None
if self.use_labels:
_A : Tuple = ids_tensor([self.batch_size] , self.num_labels )
_A : Union[str, Any] = self.get_config()
return config, pixel_values, labels
def a__ ( self ) -> List[str]:
return RegNetConfig(
num_channels=self.num_channels , embeddings_size=self.embeddings_size , hidden_sizes=self.hidden_sizes , depths=self.depths , hidden_act=self.hidden_act , num_labels=self.num_labels , )
def a__ ( self , _a , _a , _a ) -> str:
_A : List[str] = TFRegNetModel(config=_a )
_A : Optional[Any] = model(_a , training=_a )
# expected last hidden states: B, C, H // 32, W // 32
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.hidden_sizes[-1], self.image_size // 32, self.image_size // 32) , )
def a__ ( self , _a , _a , _a ) -> Dict:
_A : List[str] = self.num_labels
_A : str = TFRegNetForImageClassification(_a )
_A : List[Any] = model(_a , labels=_a , training=_a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) )
def a__ ( self ) -> Optional[Any]:
_A : int = self.prepare_config_and_inputs()
_A , _A , _A : List[str] = config_and_inputs
_A : List[Any] = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_tf
class lowercase ( UpperCamelCase__,UpperCamelCase__,unittest.TestCase ):
_a = (TFRegNetModel, TFRegNetForImageClassification) if is_tf_available() else ()
_a = (
{"feature-extraction": TFRegNetModel, "image-classification": TFRegNetForImageClassification}
if is_tf_available()
else {}
)
_a = False
_a = False
_a = False
_a = False
_a = False
def a__ ( self ) -> Optional[Any]:
_A : Optional[int] = TFRegNetModelTester(self )
_A : str = ConfigTester(self , config_class=_a , has_text_modality=_a )
def a__ ( self ) -> Union[str, Any]:
return
@unittest.skip(reason="""RegNet does not use inputs_embeds""" )
def a__ ( self ) -> Optional[int]:
pass
@unittest.skipIf(
not is_tf_available() or len(tf.config.list_physical_devices("""GPU""" ) ) == 0 , reason="""TF does not support backprop for grouped convolutions on CPU.""" , )
@slow
def a__ ( self ) -> List[Any]:
super().test_keras_fit()
@unittest.skip(reason="""RegNet does not support input and output embeddings""" )
def a__ ( self ) -> Any:
pass
def a__ ( self ) -> str:
_A , _A : Union[str, Any] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
_A : Dict = model_class(_a )
_A : List[str] = inspect.signature(model.call )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
_A : Optional[Any] = [*signature.parameters.keys()]
_A : Union[str, Any] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , _a )
def a__ ( self ) -> Dict:
_A : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*_a )
def a__ ( self ) -> Dict:
def check_hidden_states_output(_a , _a , _a ):
_A : int = model_class(_a )
_A : Optional[Any] = model(**self._prepare_for_class(_a , _a ) , training=_a )
_A : List[str] = outputs.encoder_hidden_states if config.is_encoder_decoder else outputs.hidden_states
_A : Optional[Any] = self.model_tester.num_stages
self.assertEqual(len(_a ) , expected_num_stages + 1 )
# RegNet's feature maps are of shape (batch_size, num_channels, height, width)
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [self.model_tester.image_size // 2, self.model_tester.image_size // 2] , )
_A , _A : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
_A : Dict = ["""basic""", """bottleneck"""]
for model_class in self.all_model_classes:
for layer_type in layers_type:
_A : Tuple = layer_type
_A : str = True
check_hidden_states_output(_a , _a , _a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
_A : Any = True
check_hidden_states_output(_a , _a , _a )
def a__ ( self ) -> Dict:
_A , _A : Dict = self.model_tester.prepare_config_and_inputs_for_common()
def check_equivalence(_a , _a , _a , _a={} ):
_A : Dict = model(_a , return_dict=_a , **_a )
_A : int = model(_a , return_dict=_a , **_a ).to_tuple()
def recursive_check(_a , _a ):
if isinstance(_a , (List, Tuple) ):
for tuple_iterable_value, dict_iterable_value in zip(_a , _a ):
recursive_check(_a , _a )
elif tuple_object is None:
return
else:
self.assertTrue(
all(tf.equal(_a , _a ) ) , msg=(
"""Tuple and dict output are not equal. Difference:"""
F''' {tf.math.reduce_max(tf.abs(tuple_object - dict_object ) )}'''
) , )
recursive_check(_a , _a )
for model_class in self.all_model_classes:
_A : Dict = model_class(_a )
_A : str = self._prepare_for_class(_a , _a )
_A : Dict = self._prepare_for_class(_a , _a )
check_equivalence(_a , _a , _a )
_A : Optional[Any] = self._prepare_for_class(_a , _a , return_labels=_a )
_A : List[str] = self._prepare_for_class(_a , _a , return_labels=_a )
check_equivalence(_a , _a , _a )
_A : Optional[Any] = self._prepare_for_class(_a , _a )
_A : List[Any] = self._prepare_for_class(_a , _a )
check_equivalence(_a , _a , _a , {"""output_hidden_states""": True} )
_A : int = self._prepare_for_class(_a , _a , return_labels=_a )
_A : List[Any] = self._prepare_for_class(_a , _a , return_labels=_a )
check_equivalence(_a , _a , _a , {"""output_hidden_states""": True} )
def a__ ( self ) -> List[Any]:
_A : List[Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*_a )
@slow
def a__ ( self ) -> Optional[int]:
for model_name in TF_REGNET_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
_A : List[Any] = TFRegNetModel.from_pretrained(_a )
self.assertIsNotNone(_a )
def lowerCAmelCase_ ( ):
_A : Any = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_tf
@require_vision
class lowercase ( unittest.TestCase ):
@cached_property
def a__ ( self ) -> str:
return (
AutoImageProcessor.from_pretrained(TF_REGNET_PRETRAINED_MODEL_ARCHIVE_LIST[0] )
if is_vision_available()
else None
)
@slow
def a__ ( self ) -> Dict:
_A : Optional[int] = TFRegNetForImageClassification.from_pretrained(TF_REGNET_PRETRAINED_MODEL_ARCHIVE_LIST[0] )
_A : str = self.default_image_processor
_A : Dict = prepare_img()
_A : int = image_processor(images=_a , return_tensors="""tf""" )
# forward pass
_A : Any = model(**_a , training=_a )
# verify the logits
_A : str = tf.TensorShape((1, 1000) )
self.assertEqual(outputs.logits.shape , _a )
_A : str = tf.constant([-0.4180, -1.5051, -3.4836] )
tf.debugging.assert_near(outputs.logits[0, :3] , _a , atol=1e-4 )
| 26 |
import multiprocessing
from typing import TYPE_CHECKING, Optional, Union
from .. import Dataset, Features, config
from ..formatting import query_table
from ..packaged_modules.sql.sql import Sql
from ..utils import logging
from .abc import AbstractDatasetInputStream
if TYPE_CHECKING:
import sqlitea
import sqlalchemy
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a = None , _a = None , _a = False , **_a , ) -> int:
super().__init__(features=_a , cache_dir=_a , keep_in_memory=_a , **_a )
_A : Optional[int] = Sql(
cache_dir=_a , features=_a , sql=_a , con=_a , **_a , )
def a__ ( self ) -> Optional[Any]:
_A : Tuple = None
_A : int = None
_A : Tuple = None
_A : Union[str, Any] = None
self.builder.download_and_prepare(
download_config=_a , download_mode=_a , verification_mode=_a , base_path=_a , )
# Build dataset for splits
_A : int = self.builder.as_dataset(
split="""train""" , verification_mode=_a , in_memory=self.keep_in_memory )
return dataset
class lowercase :
def __init__( self , _a , _a , _a , _a = None , _a = None , **_a , ) -> Union[str, Any]:
if num_proc is not None and num_proc <= 0:
raise ValueError(F'''num_proc {num_proc} must be an integer > 0.''' )
_A : Dict = dataset
_A : int = name
_A : Union[str, Any] = con
_A : str = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE
_A : str = num_proc
_A : Optional[Any] = to_sql_kwargs
def a__ ( self ) -> int:
_A : Any = self.to_sql_kwargs.pop("""sql""" , _a )
_A : List[str] = self.to_sql_kwargs.pop("""con""" , _a )
_A : int = self.to_sql_kwargs.pop("""index""" , _a )
_A : List[str] = self._write(index=_a , **self.to_sql_kwargs )
return written
def a__ ( self , _a ) -> Optional[int]:
_A , _A , _A : List[str] = args
_A : int = {**to_sql_kwargs, """if_exists""": """append"""} if offset > 0 else to_sql_kwargs
_A : str = query_table(
table=self.dataset.data , key=slice(_a , offset + self.batch_size ) , indices=self.dataset._indices , )
_A : Tuple = batch.to_pandas()
_A : Union[str, Any] = df.to_sql(self.name , self.con , index=_a , **_a )
return num_rows or len(_a )
def a__ ( self , _a , **_a ) -> int:
_A : Any = 0
if self.num_proc is None or self.num_proc == 1:
for offset in logging.tqdm(
range(0 , len(self.dataset ) , self.batch_size ) , unit="""ba""" , disable=not logging.is_progress_bar_enabled() , desc="""Creating SQL from Arrow format""" , ):
written += self._batch_sql((offset, index, to_sql_kwargs) )
else:
_A , _A : Tuple = len(self.dataset ), self.batch_size
with multiprocessing.Pool(self.num_proc ) as pool:
for num_rows in logging.tqdm(
pool.imap(
self._batch_sql , [(offset, index, to_sql_kwargs) for offset in range(0 , _a , _a )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit="""ba""" , disable=not logging.is_progress_bar_enabled() , desc="""Creating SQL from Arrow format""" , ):
written += num_rows
return written
| 26 | 1 |
def lowerCAmelCase_ ( ):
return [
a * b * (1000 - a - b)
for a in range(1,999 )
for b in range(snake_case_,999 )
if (a * a + b * b == (1000 - a - b) ** 2)
][0]
if __name__ == "__main__":
print(f"""{solution() = }""")
| 26 |
from ...configuration_utils import PretrainedConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/config.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/config.json"
# See all FNet models at https://huggingface.co/models?filter=fnet
}
class lowercase ( UpperCamelCase__ ):
_a = "fnet"
def __init__( self , _a=3_2000 , _a=768 , _a=12 , _a=3072 , _a="gelu_new" , _a=0.1 , _a=512 , _a=4 , _a=0.02 , _a=1e-12 , _a=False , _a=512 , _a=3 , _a=1 , _a=2 , **_a , ) -> int:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Any = vocab_size
_A : str = max_position_embeddings
_A : Optional[Any] = hidden_size
_A : List[str] = num_hidden_layers
_A : List[str] = intermediate_size
_A : List[Any] = hidden_act
_A : List[str] = hidden_dropout_prob
_A : List[str] = initializer_range
_A : List[Any] = type_vocab_size
_A : List[Any] = layer_norm_eps
_A : List[str] = use_tpu_fourier_optimizations
_A : str = tpu_short_seq_length
| 26 | 1 |
from __future__ import annotations
import numpy as np
def lowerCAmelCase_ ( snake_case_ ):
return np.maximum(0,snake_case_ )
if __name__ == "__main__":
print(np.array(relu([-1, 0, 5]))) # --> [0, 0, 5]
| 26 |
def lowerCAmelCase_ ( snake_case_ ):
if n_term == "":
return []
_A : list = []
for temp in range(int(snake_case_ ) ):
series.append(f'''1/{temp + 1}''' if series else """1""" )
return series
if __name__ == "__main__":
_snake_case = input("Enter the last number (nth term) of the Harmonic Series")
print("Formula of Harmonic Series => 1+1/2+1/3 ..... 1/n")
print(harmonic_series(nth_term))
| 26 | 1 |
from typing import Dict, Iterable, List, Optional, Union
import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
normalize,
rescale,
resize,
to_channel_dimension_format,
)
from ...image_utils import (
IMAGENET_DEFAULT_MEAN,
IMAGENET_DEFAULT_STD,
ChannelDimension,
ImageInput,
PILImageResampling,
make_list_of_images,
to_numpy_array,
valid_images,
)
from ...utils import TensorType, logging
_snake_case = logging.get_logger(__name__)
class lowercase ( UpperCamelCase__ ):
_a = ["pixel_values"]
def __init__( self , _a = True , _a = None , _a = PILImageResampling.BICUBIC , _a = True , _a = None , _a = True , _a = 1 / 255 , _a = True , _a = IMAGENET_DEFAULT_MEAN , _a = IMAGENET_DEFAULT_STD , **_a , ) -> None:
super().__init__(**_a )
_A : Any = size if size is not None else {"""shortest_edge""": 224}
_A : str = get_size_dict(_a , default_to_square=_a )
_A : List[Any] = crop_size if crop_size is not None else {"""height""": 224, """width""": 224}
_A : int = get_size_dict(_a , param_name="""crop_size""" )
_A : str = do_resize
_A : List[str] = size
_A : str = resample
_A : Dict = do_center_crop
_A : Optional[Any] = crop_size
_A : Optional[int] = do_rescale
_A : Optional[Any] = rescale_factor
_A : Tuple = do_normalize
_A : int = image_mean if image_mean is not None else IMAGENET_DEFAULT_MEAN
_A : Optional[int] = image_std if image_std is not None else IMAGENET_DEFAULT_STD
def a__ ( self , _a , _a , _a = PILImageResampling.BICUBIC , _a = None , **_a , ) -> np.ndarray:
_A : Optional[Any] = get_size_dict(_a , default_to_square=_a )
# size_dict is a dict with either keys "height" and "width" or "shortest_edge"
if "shortest_edge" in size:
_A : Optional[Any] = int((256 / 224) * size["""shortest_edge"""] )
_A : Any = get_resize_output_image_size(_a , size=_a , default_to_square=_a )
_A : str = {"""height""": output_size[0], """width""": output_size[1]}
if "height" not in size_dict or "width" not in size_dict:
raise ValueError(
F'''Size dict must have keys \'height\' and \'width\' or \'shortest_edge\'. Got {size_dict.keys()}''' )
return resize(
_a , size=(size_dict["""height"""], size_dict["""width"""]) , resample=_a , data_format=_a , **_a )
def a__ ( self , _a , _a , _a = None , **_a , ) -> np.ndarray:
_A : Optional[int] = get_size_dict(_a )
if "height" not in size or "width" not in size:
raise ValueError(F'''Size dict must have keys \'height\' and \'width\'. Got {size.keys()}''' )
return center_crop(_a , size=(size["""height"""], size["""width"""]) , data_format=_a , **_a )
def a__ ( self , _a , _a , _a = None , **_a , ) -> np.ndarray:
return rescale(_a , scale=_a , data_format=_a , **_a )
def a__ ( self , _a , _a , _a , _a = None , **_a , ) -> np.ndarray:
return normalize(_a , mean=_a , std=_a , data_format=_a , **_a )
def a__ ( self , _a , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = None , _a = ChannelDimension.FIRST , **_a , ) -> BatchFeature:
_A : Dict = do_resize if do_resize is not None else self.do_resize
_A : List[Any] = resample if resample is not None else self.resample
_A : Optional[int] = do_center_crop if do_center_crop is not None else self.do_center_crop
_A : Union[str, Any] = do_rescale if do_rescale is not None else self.do_rescale
_A : Tuple = rescale_factor if rescale_factor is not None else self.rescale_factor
_A : List[Any] = do_normalize if do_normalize is not None else self.do_normalize
_A : Optional[int] = image_mean if image_mean is not None else self.image_mean
_A : int = image_std if image_std is not None else self.image_std
_A : List[Any] = size if size is not None else self.size
_A : Any = get_size_dict(_a , default_to_square=_a )
_A : Optional[int] = crop_size if crop_size is not None else self.crop_size
_A : Optional[Any] = get_size_dict(_a , param_name="""crop_size""" )
_A : List[str] = make_list_of_images(_a )
if not valid_images(_a ):
raise ValueError(
"""Invalid image type. Must be of type PIL.Image.Image, numpy.ndarray, """
"""torch.Tensor, tf.Tensor or jax.ndarray.""" )
if do_resize and size is None:
raise ValueError("""Size must be specified if do_resize is True.""" )
if do_center_crop and crop_size is None:
raise ValueError("""Crop size must be specified if do_center_crop is True.""" )
if do_rescale and rescale_factor is None:
raise ValueError("""Rescale factor must be specified if do_rescale is True.""" )
if do_normalize and (image_mean is None or image_std is None):
raise ValueError("""Image mean and std must be specified if do_normalize is True.""" )
# All transformations expect numpy arrays.
_A : int = [to_numpy_array(_a ) for image in images]
if do_resize:
_A : Tuple = [self.resize(_a , _a , _a ) for image in images]
if do_center_crop:
_A : Dict = [self.center_crop(_a , _a ) for image in images]
if do_rescale:
_A : Dict = [self.rescale(_a , _a ) for image in images]
if do_normalize:
_A : Optional[int] = [self.normalize(_a , _a , _a ) for image in images]
_A : str = [to_channel_dimension_format(_a , _a ) for image in images]
_A : int = {"""pixel_values""": images}
return BatchFeature(data=_a , tensor_type=_a )
| 26 |
import importlib
import json
import os
from collections import OrderedDict
from typing import Dict, Optional, Union
# Build the list of all feature extractors
from ...configuration_utils import PretrainedConfig
from ...dynamic_module_utils import get_class_from_dynamic_module, resolve_trust_remote_code
from ...feature_extraction_utils import FeatureExtractionMixin
from ...utils import CONFIG_NAME, FEATURE_EXTRACTOR_NAME, get_file_from_repo, logging
from .auto_factory import _LazyAutoMapping
from .configuration_auto import (
CONFIG_MAPPING_NAMES,
AutoConfig,
model_type_to_module_name,
replace_list_option_in_docstrings,
)
_snake_case = logging.get_logger(__name__)
_snake_case = OrderedDict(
[
("audio-spectrogram-transformer", "ASTFeatureExtractor"),
("beit", "BeitFeatureExtractor"),
("chinese_clip", "ChineseCLIPFeatureExtractor"),
("clap", "ClapFeatureExtractor"),
("clip", "CLIPFeatureExtractor"),
("clipseg", "ViTFeatureExtractor"),
("conditional_detr", "ConditionalDetrFeatureExtractor"),
("convnext", "ConvNextFeatureExtractor"),
("cvt", "ConvNextFeatureExtractor"),
("data2vec-audio", "Wav2Vec2FeatureExtractor"),
("data2vec-vision", "BeitFeatureExtractor"),
("deformable_detr", "DeformableDetrFeatureExtractor"),
("deit", "DeiTFeatureExtractor"),
("detr", "DetrFeatureExtractor"),
("dinat", "ViTFeatureExtractor"),
("donut-swin", "DonutFeatureExtractor"),
("dpt", "DPTFeatureExtractor"),
("encodec", "EncodecFeatureExtractor"),
("flava", "FlavaFeatureExtractor"),
("glpn", "GLPNFeatureExtractor"),
("groupvit", "CLIPFeatureExtractor"),
("hubert", "Wav2Vec2FeatureExtractor"),
("imagegpt", "ImageGPTFeatureExtractor"),
("layoutlmv2", "LayoutLMv2FeatureExtractor"),
("layoutlmv3", "LayoutLMv3FeatureExtractor"),
("levit", "LevitFeatureExtractor"),
("maskformer", "MaskFormerFeatureExtractor"),
("mctct", "MCTCTFeatureExtractor"),
("mobilenet_v1", "MobileNetV1FeatureExtractor"),
("mobilenet_v2", "MobileNetV2FeatureExtractor"),
("mobilevit", "MobileViTFeatureExtractor"),
("nat", "ViTFeatureExtractor"),
("owlvit", "OwlViTFeatureExtractor"),
("perceiver", "PerceiverFeatureExtractor"),
("poolformer", "PoolFormerFeatureExtractor"),
("regnet", "ConvNextFeatureExtractor"),
("resnet", "ConvNextFeatureExtractor"),
("segformer", "SegformerFeatureExtractor"),
("sew", "Wav2Vec2FeatureExtractor"),
("sew-d", "Wav2Vec2FeatureExtractor"),
("speech_to_text", "Speech2TextFeatureExtractor"),
("speecht5", "SpeechT5FeatureExtractor"),
("swiftformer", "ViTFeatureExtractor"),
("swin", "ViTFeatureExtractor"),
("swinv2", "ViTFeatureExtractor"),
("table-transformer", "DetrFeatureExtractor"),
("timesformer", "VideoMAEFeatureExtractor"),
("tvlt", "TvltFeatureExtractor"),
("unispeech", "Wav2Vec2FeatureExtractor"),
("unispeech-sat", "Wav2Vec2FeatureExtractor"),
("van", "ConvNextFeatureExtractor"),
("videomae", "VideoMAEFeatureExtractor"),
("vilt", "ViltFeatureExtractor"),
("vit", "ViTFeatureExtractor"),
("vit_mae", "ViTFeatureExtractor"),
("vit_msn", "ViTFeatureExtractor"),
("wav2vec2", "Wav2Vec2FeatureExtractor"),
("wav2vec2-conformer", "Wav2Vec2FeatureExtractor"),
("wavlm", "Wav2Vec2FeatureExtractor"),
("whisper", "WhisperFeatureExtractor"),
("xclip", "CLIPFeatureExtractor"),
("yolos", "YolosFeatureExtractor"),
]
)
_snake_case = _LazyAutoMapping(CONFIG_MAPPING_NAMES, FEATURE_EXTRACTOR_MAPPING_NAMES)
def lowerCAmelCase_ ( snake_case_ ):
for module_name, extractors in FEATURE_EXTRACTOR_MAPPING_NAMES.items():
if class_name in extractors:
_A : List[str] = model_type_to_module_name(snake_case_ )
_A : List[Any] = importlib.import_module(f'''.{module_name}''',"""transformers.models""" )
try:
return getattr(snake_case_,snake_case_ )
except AttributeError:
continue
for _, extractor in FEATURE_EXTRACTOR_MAPPING._extra_content.items():
if getattr(snake_case_,"""__name__""",snake_case_ ) == class_name:
return extractor
# We did not fine the class, but maybe it's because a dep is missing. In that case, the class will be in the main
# init and we return the proper dummy to get an appropriate error message.
_A : List[Any] = importlib.import_module("""transformers""" )
if hasattr(snake_case_,snake_case_ ):
return getattr(snake_case_,snake_case_ )
return None
def lowerCAmelCase_ ( snake_case_,snake_case_ = None,snake_case_ = False,snake_case_ = False,snake_case_ = None,snake_case_ = None,snake_case_ = None,snake_case_ = False,**snake_case_,):
_A : Optional[int] = get_file_from_repo(
snake_case_,snake_case_,cache_dir=snake_case_,force_download=snake_case_,resume_download=snake_case_,proxies=snake_case_,use_auth_token=snake_case_,revision=snake_case_,local_files_only=snake_case_,)
if resolved_config_file is None:
logger.info(
"""Could not locate the feature extractor configuration file, will try to use the model config instead.""" )
return {}
with open(snake_case_,encoding="""utf-8""" ) as reader:
return json.load(snake_case_ )
class lowercase :
def __init__( self ) -> List[Any]:
raise EnvironmentError(
"""AutoFeatureExtractor is designed to be instantiated """
"""using the `AutoFeatureExtractor.from_pretrained(pretrained_model_name_or_path)` method.""" )
@classmethod
@replace_list_option_in_docstrings(_a )
def a__ ( cls , _a , **_a ) -> Any:
_A : Tuple = kwargs.pop("""config""" , _a )
_A : Tuple = kwargs.pop("""trust_remote_code""" , _a )
_A : List[Any] = True
_A , _A : Tuple = FeatureExtractionMixin.get_feature_extractor_dict(_a , **_a )
_A : Tuple = config_dict.get("""feature_extractor_type""" , _a )
_A : int = None
if "AutoFeatureExtractor" in config_dict.get("""auto_map""" , {} ):
_A : Optional[int] = config_dict["""auto_map"""]["""AutoFeatureExtractor"""]
# If we don't find the feature extractor class in the feature extractor config, let's try the model config.
if feature_extractor_class is None and feature_extractor_auto_map is None:
if not isinstance(_a , _a ):
_A : int = AutoConfig.from_pretrained(_a , **_a )
# It could be in `config.feature_extractor_type``
_A : Optional[int] = getattr(_a , """feature_extractor_type""" , _a )
if hasattr(_a , """auto_map""" ) and "AutoFeatureExtractor" in config.auto_map:
_A : Tuple = config.auto_map["""AutoFeatureExtractor"""]
if feature_extractor_class is not None:
_A : Optional[Any] = feature_extractor_class_from_name(_a )
_A : List[Any] = feature_extractor_auto_map is not None
_A : Union[str, Any] = feature_extractor_class is not None or type(_a ) in FEATURE_EXTRACTOR_MAPPING
_A : Optional[int] = resolve_trust_remote_code(
_a , _a , _a , _a )
if has_remote_code and trust_remote_code:
_A : Dict = get_class_from_dynamic_module(
_a , _a , **_a )
_A : str = kwargs.pop("""code_revision""" , _a )
if os.path.isdir(_a ):
feature_extractor_class.register_for_auto_class()
return feature_extractor_class.from_dict(_a , **_a )
elif feature_extractor_class is not None:
return feature_extractor_class.from_dict(_a , **_a )
# Last try: we use the FEATURE_EXTRACTOR_MAPPING.
elif type(_a ) in FEATURE_EXTRACTOR_MAPPING:
_A : Dict = FEATURE_EXTRACTOR_MAPPING[type(_a )]
return feature_extractor_class.from_dict(_a , **_a )
raise ValueError(
F'''Unrecognized feature extractor in {pretrained_model_name_or_path}. Should have a '''
F'''`feature_extractor_type` key in its {FEATURE_EXTRACTOR_NAME} of {CONFIG_NAME}, or one of the following '''
F'''`model_type` keys in its {CONFIG_NAME}: {", ".join(c for c in FEATURE_EXTRACTOR_MAPPING_NAMES.keys() )}''' )
@staticmethod
def a__ ( _a , _a ) -> Optional[int]:
FEATURE_EXTRACTOR_MAPPING.register(_a , _a )
| 26 | 1 |
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ = 0,snake_case_ = 0 ):
_A : Optional[int] = right or len(snake_case_ ) - 1
if left > right:
return -1
elif list_data[left] == key:
return left
elif list_data[right] == key:
return right
else:
return search(snake_case_,snake_case_,left + 1,right - 1 )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 |
import unittest
import numpy as np
from transformers.testing_utils import is_flaky, require_torch, require_vision
from transformers.utils import is_torch_available, is_vision_available
from ...test_image_processing_common import ImageProcessingSavingTestMixin, prepare_image_inputs
if is_torch_available():
import torch
if is_vision_available():
from PIL import Image
from transformers import DonutImageProcessor
class lowercase ( unittest.TestCase ):
def __init__( self , _a , _a=7 , _a=3 , _a=18 , _a=30 , _a=400 , _a=True , _a=None , _a=True , _a=False , _a=True , _a=True , _a=[0.5, 0.5, 0.5] , _a=[0.5, 0.5, 0.5] , ) -> Dict:
_A : str = parent
_A : int = batch_size
_A : Optional[int] = num_channels
_A : List[Any] = image_size
_A : int = min_resolution
_A : Optional[int] = max_resolution
_A : Any = do_resize
_A : List[str] = size if size is not None else {"""height""": 18, """width""": 20}
_A : Optional[int] = do_thumbnail
_A : str = do_align_axis
_A : List[Any] = do_pad
_A : Optional[Any] = do_normalize
_A : Tuple = image_mean
_A : List[str] = image_std
def a__ ( self ) -> Optional[int]:
return {
"do_resize": self.do_resize,
"size": self.size,
"do_thumbnail": self.do_thumbnail,
"do_align_long_axis": self.do_align_axis,
"do_pad": self.do_pad,
"do_normalize": self.do_normalize,
"image_mean": self.image_mean,
"image_std": self.image_std,
}
@require_torch
@require_vision
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = DonutImageProcessor if is_vision_available() else None
def a__ ( self ) -> Optional[int]:
_A : List[str] = DonutImageProcessingTester(self )
@property
def a__ ( self ) -> List[Any]:
return self.image_processor_tester.prepare_image_processor_dict()
def a__ ( self ) -> Optional[Any]:
_A : Union[str, Any] = self.image_processing_class(**self.image_processor_dict )
self.assertTrue(hasattr(_a , """do_resize""" ) )
self.assertTrue(hasattr(_a , """size""" ) )
self.assertTrue(hasattr(_a , """do_thumbnail""" ) )
self.assertTrue(hasattr(_a , """do_align_long_axis""" ) )
self.assertTrue(hasattr(_a , """do_pad""" ) )
self.assertTrue(hasattr(_a , """do_normalize""" ) )
self.assertTrue(hasattr(_a , """image_mean""" ) )
self.assertTrue(hasattr(_a , """image_std""" ) )
def a__ ( self ) -> List[Any]:
_A : Union[str, Any] = self.image_processing_class.from_dict(self.image_processor_dict )
self.assertEqual(image_processor.size , {"""height""": 18, """width""": 20} )
_A : Optional[int] = self.image_processing_class.from_dict(self.image_processor_dict , size=42 )
self.assertEqual(image_processor.size , {"""height""": 42, """width""": 42} )
# Previous config had dimensions in (width, height) order
_A : List[str] = self.image_processing_class.from_dict(self.image_processor_dict , size=(42, 84) )
self.assertEqual(image_processor.size , {"""height""": 84, """width""": 42} )
def a__ ( self ) -> Union[str, Any]:
pass
@is_flaky()
def a__ ( self ) -> Optional[int]:
# Initialize image_processing
_A : List[Any] = self.image_processing_class(**self.image_processor_dict )
# create random PIL images
_A : Dict = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a )
for image in image_inputs:
self.assertIsInstance(_a , Image.Image )
# Test not batched input
_A : Union[str, Any] = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : Any = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
@is_flaky()
def a__ ( self ) -> Dict:
# Initialize image_processing
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random numpy tensors
_A : Tuple = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , numpify=_a )
for image in image_inputs:
self.assertIsInstance(_a , np.ndarray )
# Test not batched input
_A : int = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : List[str] = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
@is_flaky()
def a__ ( self ) -> Optional[int]:
# Initialize image_processing
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random PyTorch tensors
_A : Optional[int] = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , torchify=_a )
for image in image_inputs:
self.assertIsInstance(_a , torch.Tensor )
# Test not batched input
_A : Any = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : str = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
| 26 | 1 |
import warnings
from typing import List, Optional, Union
from ...processing_utils import ProcessorMixin
from ...tokenization_utils_base import BatchEncoding, PaddingStrategy, PreTokenizedInput, TextInput, TruncationStrategy
from ...utils import TensorType
class lowercase ( UpperCamelCase__ ):
_a = ["image_processor", "tokenizer"]
_a = "LayoutLMv3ImageProcessor"
_a = ("LayoutLMv3Tokenizer", "LayoutLMv3TokenizerFast")
def __init__( self , _a=None , _a=None , **_a ) -> Tuple:
_A : Union[str, Any] = None
if "feature_extractor" in kwargs:
warnings.warn(
"""The `feature_extractor` argument is deprecated and will be removed in v5, use `image_processor`"""
""" instead.""" , _a , )
_A : Optional[int] = kwargs.pop("""feature_extractor""" )
_A : Union[str, Any] = image_processor if image_processor is not None else feature_extractor
if image_processor is None:
raise ValueError("""You need to specify an `image_processor`.""" )
if tokenizer is None:
raise ValueError("""You need to specify a `tokenizer`.""" )
super().__init__(_a , _a )
def __call__( self , _a , _a = None , _a = None , _a = None , _a = None , _a = True , _a = False , _a = None , _a = None , _a = 0 , _a = None , _a = None , _a = None , _a = False , _a = False , _a = False , _a = False , _a = True , _a = None , **_a , ) -> BatchEncoding:
# verify input
if self.image_processor.apply_ocr and (boxes is not None):
raise ValueError(
"""You cannot provide bounding boxes if you initialized the image processor with apply_ocr set to True.""" )
if self.image_processor.apply_ocr and (word_labels is not None):
raise ValueError(
"""You cannot provide word labels if you initialized the image processor with apply_ocr set to True.""" )
# first, apply the image processor
_A : Dict = self.image_processor(images=_a , return_tensors=_a )
# second, apply the tokenizer
if text is not None and self.image_processor.apply_ocr and text_pair is None:
if isinstance(_a , _a ):
_A : Tuple = [text] # add batch dimension (as the image processor always adds a batch dimension)
_A : Any = features["""words"""]
_A : Optional[Any] = self.tokenizer(
text=text if text is not None else features["""words"""] , text_pair=text_pair if text_pair is not None else None , boxes=boxes if boxes is not None else features["""boxes"""] , word_labels=_a , add_special_tokens=_a , padding=_a , truncation=_a , max_length=_a , stride=_a , pad_to_multiple_of=_a , return_token_type_ids=_a , return_attention_mask=_a , return_overflowing_tokens=_a , return_special_tokens_mask=_a , return_offsets_mapping=_a , return_length=_a , verbose=_a , return_tensors=_a , **_a , )
# add pixel values
_A : Tuple = features.pop("""pixel_values""" )
if return_overflowing_tokens is True:
_A : Union[str, Any] = self.get_overflowing_images(_a , encoded_inputs["""overflow_to_sample_mapping"""] )
_A : Union[str, Any] = images
return encoded_inputs
def a__ ( self , _a , _a ) -> Optional[int]:
# in case there's an overflow, ensure each `input_ids` sample is mapped to its corresponding image
_A : List[str] = []
for sample_idx in overflow_to_sample_mapping:
images_with_overflow.append(images[sample_idx] )
if len(_a ) != len(_a ):
raise ValueError(
"""Expected length of images to be the same as the length of `overflow_to_sample_mapping`, but got"""
F''' {len(_a )} and {len(_a )}''' )
return images_with_overflow
def a__ ( self , *_a , **_a ) -> Optional[int]:
return self.tokenizer.batch_decode(*_a , **_a )
def a__ ( self , *_a , **_a ) -> List[Any]:
return self.tokenizer.decode(*_a , **_a )
@property
def a__ ( self ) -> List[Any]:
return ["input_ids", "bbox", "attention_mask", "pixel_values"]
@property
def a__ ( self ) -> Union[str, Any]:
warnings.warn(
"""`feature_extractor_class` is deprecated and will be removed in v5. Use `image_processor_class` instead.""" , _a , )
return self.image_processor_class
@property
def a__ ( self ) -> int:
warnings.warn(
"""`feature_extractor` is deprecated and will be removed in v5. Use `image_processor` instead.""" , _a , )
return self.image_processor
| 26 |
from __future__ import annotations
import numpy as np
def lowerCAmelCase_ ( snake_case_ ):
return np.maximum(0,snake_case_ )
if __name__ == "__main__":
print(np.array(relu([-1, 0, 5]))) # --> [0, 0, 5]
| 26 | 1 |
import torch
from diffusers import UnCLIPScheduler
from .test_schedulers import SchedulerCommonTest
class lowercase ( UpperCamelCase__ ):
_a = (UnCLIPScheduler,)
def a__ ( self , **_a ) -> Any:
_A : Dict = {
"""num_train_timesteps""": 1000,
"""variance_type""": """fixed_small_log""",
"""clip_sample""": True,
"""clip_sample_range""": 1.0,
"""prediction_type""": """epsilon""",
}
config.update(**_a )
return config
def a__ ( self ) -> Union[str, Any]:
for timesteps in [1, 5, 100, 1000]:
self.check_over_configs(num_train_timesteps=_a )
def a__ ( self ) -> str:
for variance in ["fixed_small_log", "learned_range"]:
self.check_over_configs(variance_type=_a )
def a__ ( self ) -> List[str]:
for clip_sample in [True, False]:
self.check_over_configs(clip_sample=_a )
def a__ ( self ) -> Optional[Any]:
for clip_sample_range in [1, 5, 10, 20]:
self.check_over_configs(clip_sample_range=_a )
def a__ ( self ) -> str:
for prediction_type in ["epsilon", "sample"]:
self.check_over_configs(prediction_type=_a )
def a__ ( self ) -> Tuple:
for time_step in [0, 500, 999]:
for prev_timestep in [None, 5, 100, 250, 500, 750]:
if prev_timestep is not None and prev_timestep >= time_step:
continue
self.check_over_forward(time_step=_a , prev_timestep=_a )
def a__ ( self ) -> str:
_A : str = self.scheduler_classes[0]
_A : str = self.get_scheduler_config(variance_type="""fixed_small_log""" )
_A : Optional[Any] = scheduler_class(**_a )
assert torch.sum(torch.abs(scheduler._get_variance(0 ) - 1.00_00e-10 ) ) < 1e-5
assert torch.sum(torch.abs(scheduler._get_variance(487 ) - 0.0549625 ) ) < 1e-5
assert torch.sum(torch.abs(scheduler._get_variance(999 ) - 0.9994987 ) ) < 1e-5
def a__ ( self ) -> int:
_A : Dict = self.scheduler_classes[0]
_A : List[str] = self.get_scheduler_config(variance_type="""learned_range""" )
_A : Optional[Any] = scheduler_class(**_a )
_A : str = 0.5
assert scheduler._get_variance(1 , predicted_variance=_a ) - -10.1712790 < 1e-5
assert scheduler._get_variance(487 , predicted_variance=_a ) - -5.7998052 < 1e-5
assert scheduler._get_variance(999 , predicted_variance=_a ) - -0.0010011 < 1e-5
def a__ ( self ) -> Any:
_A : Any = self.scheduler_classes[0]
_A : Tuple = self.get_scheduler_config()
_A : Optional[Any] = scheduler_class(**_a )
_A : Optional[int] = scheduler.timesteps
_A : Union[str, Any] = self.dummy_model()
_A : Any = self.dummy_sample_deter
_A : List[str] = torch.manual_seed(0 )
for i, t in enumerate(_a ):
# 1. predict noise residual
_A : Dict = model(_a , _a )
# 2. predict previous mean of sample x_t-1
_A : List[Any] = scheduler.step(_a , _a , _a , generator=_a ).prev_sample
_A : Any = pred_prev_sample
_A : int = torch.sum(torch.abs(_a ) )
_A : Optional[int] = torch.mean(torch.abs(_a ) )
assert abs(result_sum.item() - 252.2682495 ) < 1e-2
assert abs(result_mean.item() - 0.3284743 ) < 1e-3
def a__ ( self ) -> Any:
_A : Dict = self.scheduler_classes[0]
_A : Optional[int] = self.get_scheduler_config()
_A : int = scheduler_class(**_a )
scheduler.set_timesteps(25 )
_A : List[str] = scheduler.timesteps
_A : str = self.dummy_model()
_A : Union[str, Any] = self.dummy_sample_deter
_A : str = torch.manual_seed(0 )
for i, t in enumerate(_a ):
# 1. predict noise residual
_A : List[Any] = model(_a , _a )
if i + 1 == timesteps.shape[0]:
_A : Dict = None
else:
_A : Optional[Any] = timesteps[i + 1]
# 2. predict previous mean of sample x_t-1
_A : Any = scheduler.step(
_a , _a , _a , prev_timestep=_a , generator=_a ).prev_sample
_A : int = pred_prev_sample
_A : Optional[int] = torch.sum(torch.abs(_a ) )
_A : int = torch.mean(torch.abs(_a ) )
assert abs(result_sum.item() - 258.2044983 ) < 1e-2
assert abs(result_mean.item() - 0.3362038 ) < 1e-3
def a__ ( self ) -> Dict:
pass
def a__ ( self ) -> Optional[Any]:
pass
| 26 |
import argparse
import shutil
import time
from json import JSONDecodeError
from logging import getLogger
from pathlib import Path
from typing import Dict, List
import torch
from torch.utils.data import DataLoader
from tqdm import tqdm
from transformers import AutoModelForSeqaSeqLM, AutoTokenizer
from utils import (
SeqaSeqDataset,
calculate_bleu,
calculate_rouge,
chunks,
lmap,
load_json,
parse_numeric_n_bool_cl_kwargs,
save_json,
use_task_specific_params,
write_txt_file,
)
_snake_case = getLogger(__name__)
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ = 8,snake_case_ = 1024,snake_case_="val",snake_case_=None,snake_case_=False,snake_case_="summarization",snake_case_=None,snake_case_=1,snake_case_ = None,snake_case_="",**snake_case_,):
_A : Dict = str(snake_case_ )
assert local_rank is not None
torch.distributed.init_process_group(backend="""nccl""",rank=snake_case_ )
_A : Tuple = Path(snake_case_ )
_A : List[Any] = save_dir.joinpath(f'''rank_{local_rank}_output.json''' )
torch.cuda.set_device(snake_case_ )
_A : Union[str, Any] = AutoModelForSeqaSeqLM.from_pretrained(snake_case_ ).cuda()
if fpaa:
_A : Any = model.half()
# determine if we need to increase num_beams
use_task_specific_params(snake_case_,snake_case_ ) # update config with task specific params
_A : str = generate_kwargs.pop("""num_beams""",model.config.num_beams ) # AttributeError risk?
if num_return_sequences > num_beams:
_A : int = num_return_sequences
_A : Optional[Any] = AutoTokenizer.from_pretrained(snake_case_ )
logger.info(f'''Inferred tokenizer type: {tokenizer.__class__}''' ) # if this is wrong, check config.model_type.
if max_source_length is None:
_A : Optional[int] = tokenizer.model_max_length
if prefix is None:
_A : Tuple = prefix or getattr(model.config,"""prefix""","""""" ) or """"""
_A : Optional[int] = SeqaSeqDataset(
snake_case_,snake_case_,snake_case_,max_target_length=1024,type_path=snake_case_,n_obs=snake_case_,prefix=snake_case_,**snake_case_,)
# I set shuffle=True for a more accurate progress bar.
# If all the longest samples are first, the prog bar estimate is too high at the beginning.
_A : Optional[int] = ds.make_sortish_sampler(snake_case_,distributed=snake_case_,add_extra_examples=snake_case_,shuffle=snake_case_ )
_A : Dict = DataLoader(snake_case_,sampler=snake_case_,batch_size=snake_case_,collate_fn=ds.collate_fn )
_A : Optional[Any] = []
for batch in tqdm(snake_case_ ):
_A : Tuple = model.generate(
input_ids=batch["""input_ids"""].to(model.device ),attention_mask=batch["""attention_mask"""].to(model.device ),num_return_sequences=snake_case_,num_beams=snake_case_,**snake_case_,)
_A : Any = tokenizer.batch_decode(snake_case_,skip_special_tokens=snake_case_,clean_up_tokenization_spaces=snake_case_ )
_A : Dict = batch["""ids"""]
if num_return_sequences > 1:
_A : Any = chunks(snake_case_,snake_case_ ) # batch size chunks, each of size num_return_seq
for i, pred in enumerate(snake_case_ ):
results.append({"""pred""": pred, """id""": ids[i].item()} )
save_json(snake_case_,snake_case_ )
return results, sampler.num_replicas
def lowerCAmelCase_ ( ):
_A : Tuple = argparse.ArgumentParser(
epilog="""Unspecified args like --num_beams=2 --decoder_start_token_id=4 are passed to model.generate""" )
parser.add_argument("""--data_dir""",type=snake_case_,help="""like cnn_dm/test.source""" )
parser.add_argument(
"""--model_name""",type=snake_case_,help="""like facebook/bart-large-cnn,t5-base, etc.""",default="""sshleifer/distilbart-xsum-12-3""",)
parser.add_argument("""--save_dir""",type=snake_case_,help="""where to save""",default="""tmp_gen""" )
parser.add_argument("""--max_source_length""",type=snake_case_,default=snake_case_ )
parser.add_argument(
"""--type_path""",type=snake_case_,default="""test""",help="""which subset to evaluate typically train/val/test""" )
parser.add_argument("""--task""",type=snake_case_,default="""summarization""",help="""used for task_specific_params + metrics""" )
parser.add_argument("""--bs""",type=snake_case_,default=8,required=snake_case_,help="""batch size""" )
parser.add_argument(
"""--local_rank""",type=snake_case_,default=-1,required=snake_case_,help="""should be passed by distributed.launch""" )
parser.add_argument(
"""--n_obs""",type=snake_case_,default=snake_case_,required=snake_case_,help="""How many observations. Defaults to all.""" )
parser.add_argument(
"""--num_return_sequences""",type=snake_case_,default=1,required=snake_case_,help="""How many sequences to return""" )
parser.add_argument(
"""--sync_timeout""",type=snake_case_,default=600,required=snake_case_,help="""How long should master process wait for other processes to finish.""",)
parser.add_argument("""--src_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument("""--tgt_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument(
"""--prefix""",type=snake_case_,required=snake_case_,default=snake_case_,help="""will be added to the begininng of src examples""" )
parser.add_argument("""--fp16""",action="""store_true""" )
parser.add_argument("""--debug""",action="""store_true""" )
_A : Union[str, Any] = time.time()
_A , _A : List[str] = parser.parse_known_args()
_A : List[str] = parse_numeric_n_bool_cl_kwargs(snake_case_ )
if generate_kwargs and args.local_rank <= 0:
print(f'''parsed the following generate kwargs: {generate_kwargs}''' )
_A : Dict = Path(args.save_dir + """_tmp""" )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ ) # this handles locking.
_A : int = list(json_save_dir.glob("""rank_*.json""" ) )
if intermediate_files:
raise ValueError(f'''Found files at {json_save_dir} please move or remove them.''' )
# In theory, a node could finish and save before another node hits this. If this happens, we can address later.
_A : Any = {}
if args.src_lang is not None:
_A : int = args.src_lang
if args.tgt_lang is not None:
_A : Dict = args.tgt_lang
Path(args.save_dir ).mkdir(exist_ok=snake_case_ )
_A , _A : str = eval_data_dir(
args.data_dir,snake_case_,args.model_name,type_path=args.type_path,bs=args.bs,fpaa=args.fpaa,task=args.task,local_rank=args.local_rank,n_obs=args.n_obs,max_source_length=args.max_source_length,num_return_sequences=args.num_return_sequences,prefix=args.prefix,dataset_kwargs=snake_case_,**snake_case_,)
if args.local_rank <= 0:
_A : List[Any] = Path(args.save_dir )
save_dir.mkdir(exist_ok=snake_case_ )
_A : Tuple = gather_results_from_each_node(snake_case_,snake_case_,args.sync_timeout )
_A : Optional[int] = combine_partial_results(snake_case_ )
if args.num_return_sequences > 1:
_A : Optional[Any] = save_dir.joinpath("""pseudolabel_results.json""" )
print(f'''Saving aggregated results at {save_path}, intermediate in {json_save_dir}/''' )
save_json(snake_case_,snake_case_ )
return
_A : List[str] = Path(args.data_dir ).joinpath(args.type_path + """.target""" )
with open(snake_case_ ) as f:
_A : int = [x.rstrip() for x in f.readlines()][: len(snake_case_ )]
# Calculate metrics, save metrics, and save _generations.txt
_A : Dict = """translation""" in args.task
_A : Optional[Any] = calculate_bleu if calc_bleu else calculate_rouge
_A : Tuple = """bleu""" if calc_bleu else """rouge"""
_A : Dict = score_fn(snake_case_,snake_case_ )
_A : List[Any] = len(snake_case_ )
_A : Optional[int] = time.time() - start_time
_A : Dict = round(runtime / metrics["""n_obs"""],4 )
_A : Dict = num_replicas
# TODO(@stas00): add whatever metadata to metrics
_A : Any = save_dir.joinpath(f'''{args.type_path}_{metric_name}.json''' )
save_json(snake_case_,snake_case_,indent=snake_case_ )
print(snake_case_ )
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}_generations.txt''' ) )
if args.debug:
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}.target''' ) )
else:
shutil.rmtree(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
_A : Dict = []
for partial_result in partial_results:
records.extend(snake_case_ )
_A : Optional[Any] = sorted(snake_case_,key=lambda snake_case_ : x["id"] )
_A : List[str] = [x["""pred"""] for x in records]
return preds
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
# WAIT FOR lots of .json files
_A : Optional[Any] = time.time()
logger.info("""waiting for all nodes to finish""" )
_A : List[str] = None
while (time.time() - start_wait) < timeout:
_A : str = list(save_dir.glob("""rank_*.json""" ) )
if len(snake_case_ ) < num_replicas:
continue
try:
# make sure all json files are fully saved
_A : List[str] = lmap(snake_case_,snake_case_ )
return json_data
except JSONDecodeError:
continue
else:
raise TimeoutError("""Rank 0 gave up on waiting for other processes""" )
# Unreachable
if __name__ == "__main__":
# Usage for MT:
run_generate()
| 26 | 1 |
import torch
def lowerCAmelCase_ ( ):
if torch.cuda.is_available():
_A : Dict = torch.cuda.device_count()
else:
_A : Union[str, Any] = 0
print(f'''Successfully ran on {num_gpus} GPUs''' )
if __name__ == "__main__":
main()
| 26 |
from __future__ import annotations
import unittest
from transformers import is_tf_available
from transformers.testing_utils import require_sentencepiece, require_tf, require_tokenizers, slow
if is_tf_available():
import numpy as np
import tensorflow as tf
from transformers import TFCamembertModel
@require_tf
@require_sentencepiece
@require_tokenizers
class lowercase ( unittest.TestCase ):
@slow
def a__ ( self ) -> Any:
_A : Tuple = TFCamembertModel.from_pretrained("""jplu/tf-camembert-base""" )
_A : List[Any] = tf.convert_to_tensor(
[[5, 121, 11, 660, 16, 730, 2_5543, 110, 83, 6]] , dtype=tf.intaa , ) # J'aime le camembert !"
_A : List[str] = model(_a )["""last_hidden_state"""]
_A : Union[str, Any] = tf.TensorShape((1, 10, 768) )
self.assertEqual(output.shape , _a )
# compare the actual values for a slice.
_A : List[Any] = tf.convert_to_tensor(
[[[-0.0254, 0.0235, 0.1027], [0.0606, -0.1811, -0.0418], [-0.1561, -0.1127, 0.2687]]] , dtype=tf.floataa , )
# camembert = torch.hub.load('pytorch/fairseq', 'camembert.v0')
# camembert.eval()
# expected_slice = roberta.model.forward(input_ids)[0][:, :3, :3].detach()
self.assertTrue(np.allclose(output[:, :3, :3].numpy() , expected_slice.numpy() , atol=1e-4 ) )
| 26 | 1 |
from typing import List, Optional, Union
import torch
from ...models import UNetaDConditionModel, VQModel
from ...pipelines import DiffusionPipeline
from ...pipelines.pipeline_utils import ImagePipelineOutput
from ...schedulers import DDPMScheduler
from ...utils import (
is_accelerate_available,
is_accelerate_version,
logging,
randn_tensor,
replace_example_docstring,
)
_snake_case = logging.get_logger(__name__) # pylint: disable=invalid-name
_snake_case = "\n Examples:\n ```py\n >>> import torch\n >>> import numpy as np\n\n >>> from diffusers import KandinskyV22PriorPipeline, KandinskyV22ControlnetPipeline\n >>> from transformers import pipeline\n >>> from diffusers.utils import load_image\n\n\n >>> def make_hint(image, depth_estimator):\n ... image = depth_estimator(image)[\"depth\"]\n ... image = np.array(image)\n ... image = image[:, :, None]\n ... image = np.concatenate([image, image, image], axis=2)\n ... detected_map = torch.from_numpy(image).float() / 255.0\n ... hint = detected_map.permute(2, 0, 1)\n ... return hint\n\n\n >>> depth_estimator = pipeline(\"depth-estimation\")\n\n >>> pipe_prior = KandinskyV22PriorPipeline.from_pretrained(\n ... \"kandinsky-community/kandinsky-2-2-prior\", torch_dtype=torch.float16\n ... )\n >>> pipe_prior = pipe_prior.to(\"cuda\")\n\n >>> pipe = KandinskyV22ControlnetPipeline.from_pretrained(\n ... \"kandinsky-community/kandinsky-2-2-controlnet-depth\", torch_dtype=torch.float16\n ... )\n >>> pipe = pipe.to(\"cuda\")\n\n\n >>> img = load_image(\n ... \"https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main\"\n ... \"/kandinsky/cat.png\"\n ... ).resize((768, 768))\n\n >>> hint = make_hint(img, depth_estimator).unsqueeze(0).half().to(\"cuda\")\n\n >>> prompt = \"A robot, 4k photo\"\n >>> negative_prior_prompt = \"lowres, text, error, cropped, worst quality, low quality, jpeg artifacts, ugly, duplicate, morbid, mutilated, out of frame, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck, username, watermark, signature\"\n\n >>> generator = torch.Generator(device=\"cuda\").manual_seed(43)\n\n >>> image_emb, zero_image_emb = pipe_prior(\n ... prompt=prompt, negative_prompt=negative_prior_prompt, generator=generator\n ... ).to_tuple()\n\n >>> images = pipe(\n ... image_embeds=image_emb,\n ... negative_image_embeds=zero_image_emb,\n ... hint=hint,\n ... num_inference_steps=50,\n ... generator=generator,\n ... height=768,\n ... width=768,\n ... ).images\n\n >>> images[0].save(\"robot_cat.png\")\n ```\n"
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_=8 ):
_A : Tuple = height // scale_factor**2
if height % scale_factor**2 != 0:
new_height += 1
_A : Union[str, Any] = width // scale_factor**2
if width % scale_factor**2 != 0:
new_width += 1
return new_height * scale_factor, new_width * scale_factor
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a , ) -> List[str]:
super().__init__()
self.register_modules(
unet=_a , scheduler=_a , movq=_a , )
_A : Tuple = 2 ** (len(self.movq.config.block_out_channels ) - 1)
def a__ ( self , _a , _a , _a , _a , _a , _a ) -> List[str]:
if latents is None:
_A : Union[str, Any] = randn_tensor(_a , generator=_a , device=_a , dtype=_a )
else:
if latents.shape != shape:
raise ValueError(F'''Unexpected latents shape, got {latents.shape}, expected {shape}''' )
_A : Tuple = latents.to(_a )
_A : Union[str, Any] = latents * scheduler.init_noise_sigma
return latents
def a__ ( self , _a=0 ) -> str:
if is_accelerate_available():
from accelerate import cpu_offload
else:
raise ImportError("""Please install accelerate via `pip install accelerate`""" )
_A : Optional[int] = torch.device(F'''cuda:{gpu_id}''' )
_A : int = [
self.unet,
self.movq,
]
for cpu_offloaded_model in models:
if cpu_offloaded_model is not None:
cpu_offload(_a , _a )
def a__ ( self , _a=0 ) -> Any:
if is_accelerate_available() and is_accelerate_version(""">=""" , """0.17.0.dev0""" ):
from accelerate import cpu_offload_with_hook
else:
raise ImportError("""`enable_model_cpu_offload` requires `accelerate v0.17.0` or higher.""" )
_A : str = torch.device(F'''cuda:{gpu_id}''' )
if self.device.type != "cpu":
self.to("""cpu""" , silence_dtype_warnings=_a )
torch.cuda.empty_cache() # otherwise we don't see the memory savings (but they probably exist)
_A : Optional[Any] = None
for cpu_offloaded_model in [self.unet, self.movq]:
_A , _A : List[str] = cpu_offload_with_hook(_a , _a , prev_module_hook=_a )
# We'll offload the last model manually.
_A : int = hook
@property
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline._execution_device
def a__ ( self ) -> str:
if not hasattr(self.unet , """_hf_hook""" ):
return self.device
for module in self.unet.modules():
if (
hasattr(_a , """_hf_hook""" )
and hasattr(module._hf_hook , """execution_device""" )
and module._hf_hook.execution_device is not None
):
return torch.device(module._hf_hook.execution_device )
return self.device
@torch.no_grad()
@replace_example_docstring(_a )
def __call__( self , _a , _a , _a , _a = 512 , _a = 512 , _a = 100 , _a = 4.0 , _a = 1 , _a = None , _a = None , _a = "pil" , _a = True , ) -> Optional[Any]:
_A : str = self._execution_device
_A : Optional[Any] = guidance_scale > 1.0
if isinstance(_a , _a ):
_A : List[Any] = torch.cat(_a , dim=0 )
if isinstance(_a , _a ):
_A : Optional[Any] = torch.cat(_a , dim=0 )
if isinstance(_a , _a ):
_A : Union[str, Any] = torch.cat(_a , dim=0 )
_A : Union[str, Any] = image_embeds.shape[0] * num_images_per_prompt
if do_classifier_free_guidance:
_A : str = image_embeds.repeat_interleave(_a , dim=0 )
_A : Tuple = negative_image_embeds.repeat_interleave(_a , dim=0 )
_A : Union[str, Any] = hint.repeat_interleave(_a , dim=0 )
_A : Optional[int] = torch.cat([negative_image_embeds, image_embeds] , dim=0 ).to(dtype=self.unet.dtype , device=_a )
_A : Optional[int] = torch.cat([hint, hint] , dim=0 ).to(dtype=self.unet.dtype , device=_a )
self.scheduler.set_timesteps(_a , device=_a )
_A : Optional[Any] = self.scheduler.timesteps
_A : str = self.movq.config.latent_channels
_A , _A : Dict = downscale_height_and_width(_a , _a , self.movq_scale_factor )
# create initial latent
_A : Any = self.prepare_latents(
(batch_size, num_channels_latents, height, width) , image_embeds.dtype , _a , _a , _a , self.scheduler , )
for i, t in enumerate(self.progress_bar(_a ) ):
# expand the latents if we are doing classifier free guidance
_A : Optional[Any] = torch.cat([latents] * 2 ) if do_classifier_free_guidance else latents
_A : int = {"""image_embeds""": image_embeds, """hint""": hint}
_A : List[Any] = self.unet(
sample=_a , timestep=_a , encoder_hidden_states=_a , added_cond_kwargs=_a , return_dict=_a , )[0]
if do_classifier_free_guidance:
_A , _A : Optional[Any] = noise_pred.split(latents.shape[1] , dim=1 )
_A , _A : Optional[int] = noise_pred.chunk(2 )
_A , _A : Optional[Any] = variance_pred.chunk(2 )
_A : Optional[Any] = noise_pred_uncond + guidance_scale * (noise_pred_text - noise_pred_uncond)
_A : List[Any] = torch.cat([noise_pred, variance_pred_text] , dim=1 )
if not (
hasattr(self.scheduler.config , """variance_type""" )
and self.scheduler.config.variance_type in ["learned", "learned_range"]
):
_A , _A : Union[str, Any] = noise_pred.split(latents.shape[1] , dim=1 )
# compute the previous noisy sample x_t -> x_t-1
_A : List[Any] = self.scheduler.step(
_a , _a , _a , generator=_a , )[0]
# post-processing
_A : Dict = self.movq.decode(_a , force_not_quantize=_a )["""sample"""]
if output_type not in ["pt", "np", "pil"]:
raise ValueError(F'''Only the output types `pt`, `pil` and `np` are supported not output_type={output_type}''' )
if output_type in ["np", "pil"]:
_A : Any = image * 0.5 + 0.5
_A : int = image.clamp(0 , 1 )
_A : Optional[int] = image.cpu().permute(0 , 2 , 3 , 1 ).float().numpy()
if output_type == "pil":
_A : Any = self.numpy_to_pil(_a )
if not return_dict:
return (image,)
return ImagePipelineOutput(images=_a )
| 26 |
from dataclasses import dataclass
from typing import List, Optional, Union
import numpy as np
import PIL
import torch
from transformers import CLIPImageProcessor, CLIPVisionModel
from ...models import PriorTransformer
from ...pipelines import DiffusionPipeline
from ...schedulers import HeunDiscreteScheduler
from ...utils import (
BaseOutput,
is_accelerate_available,
logging,
randn_tensor,
replace_example_docstring,
)
from .renderer import ShapERenderer
_snake_case = logging.get_logger(__name__) # pylint: disable=invalid-name
_snake_case = "\n Examples:\n ```py\n >>> from PIL import Image\n >>> import torch\n >>> from diffusers import DiffusionPipeline\n >>> from diffusers.utils import export_to_gif, load_image\n\n >>> device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n\n >>> repo = \"openai/shap-e-img2img\"\n >>> pipe = DiffusionPipeline.from_pretrained(repo, torch_dtype=torch.float16)\n >>> pipe = pipe.to(device)\n\n >>> guidance_scale = 3.0\n >>> image_url = \"https://hf.co/datasets/diffusers/docs-images/resolve/main/shap-e/corgi.png\"\n >>> image = load_image(image_url).convert(\"RGB\")\n\n >>> images = pipe(\n ... image,\n ... guidance_scale=guidance_scale,\n ... num_inference_steps=64,\n ... frame_size=256,\n ... ).images\n\n >>> gif_path = export_to_gif(images[0], \"corgi_3d.gif\")\n ```\n"
@dataclass
class lowercase ( UpperCamelCase__ ):
_a = 42
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a , _a , _a , ) -> List[Any]:
super().__init__()
self.register_modules(
prior=_a , image_encoder=_a , image_processor=_a , scheduler=_a , renderer=_a , )
def a__ ( self , _a , _a , _a , _a , _a , _a ) -> str:
if latents is None:
_A : str = randn_tensor(_a , generator=_a , device=_a , dtype=_a )
else:
if latents.shape != shape:
raise ValueError(F'''Unexpected latents shape, got {latents.shape}, expected {shape}''' )
_A : Union[str, Any] = latents.to(_a )
_A : int = latents * scheduler.init_noise_sigma
return latents
def a__ ( self , _a=0 ) -> Optional[Any]:
if is_accelerate_available():
from accelerate import cpu_offload
else:
raise ImportError("""Please install accelerate via `pip install accelerate`""" )
_A : str = torch.device(F'''cuda:{gpu_id}''' )
_A : Any = [self.image_encoder, self.prior]
for cpu_offloaded_model in models:
if cpu_offloaded_model is not None:
cpu_offload(_a , _a )
@property
def a__ ( self ) -> List[Any]:
if self.device != torch.device("""meta""" ) or not hasattr(self.image_encoder , """_hf_hook""" ):
return self.device
for module in self.image_encoder.modules():
if (
hasattr(_a , """_hf_hook""" )
and hasattr(module._hf_hook , """execution_device""" )
and module._hf_hook.execution_device is not None
):
return torch.device(module._hf_hook.execution_device )
return self.device
def a__ ( self , _a , _a , _a , _a , ) -> Tuple:
if isinstance(_a , _a ) and isinstance(image[0] , torch.Tensor ):
_A : int = torch.cat(_a , axis=0 ) if image[0].ndim == 4 else torch.stack(_a , axis=0 )
if not isinstance(_a , torch.Tensor ):
_A : Dict = self.image_processor(_a , return_tensors="""pt""" ).pixel_values[0].unsqueeze(0 )
_A : int = image.to(dtype=self.image_encoder.dtype , device=_a )
_A : List[Any] = self.image_encoder(_a )["""last_hidden_state"""]
_A : List[Any] = image_embeds[:, 1:, :].contiguous() # batch_size, dim, 256
_A : Dict = image_embeds.repeat_interleave(_a , dim=0 )
if do_classifier_free_guidance:
_A : str = torch.zeros_like(_a )
# For classifier free guidance, we need to do two forward passes.
# Here we concatenate the unconditional and text embeddings into a single batch
# to avoid doing two forward passes
_A : List[str] = torch.cat([negative_image_embeds, image_embeds] )
return image_embeds
@torch.no_grad()
@replace_example_docstring(_a )
def __call__( self , _a , _a = 1 , _a = 25 , _a = None , _a = None , _a = 4.0 , _a = 64 , _a = "pil" , _a = True , ) -> Union[str, Any]:
if isinstance(_a , PIL.Image.Image ):
_A : List[Any] = 1
elif isinstance(_a , torch.Tensor ):
_A : Any = image.shape[0]
elif isinstance(_a , _a ) and isinstance(image[0] , (torch.Tensor, PIL.Image.Image) ):
_A : Union[str, Any] = len(_a )
else:
raise ValueError(
F'''`image` has to be of type `PIL.Image.Image`, `torch.Tensor`, `List[PIL.Image.Image]` or `List[torch.Tensor]` but is {type(_a )}''' )
_A : Optional[int] = self._execution_device
_A : Tuple = batch_size * num_images_per_prompt
_A : List[Any] = guidance_scale > 1.0
_A : Optional[Any] = self._encode_image(_a , _a , _a , _a )
# prior
self.scheduler.set_timesteps(_a , device=_a )
_A : Optional[int] = self.scheduler.timesteps
_A : List[str] = self.prior.config.num_embeddings
_A : int = self.prior.config.embedding_dim
_A : Optional[Any] = self.prepare_latents(
(batch_size, num_embeddings * embedding_dim) , image_embeds.dtype , _a , _a , _a , self.scheduler , )
# YiYi notes: for testing only to match ldm, we can directly create a latents with desired shape: batch_size, num_embeddings, embedding_dim
_A : List[Any] = latents.reshape(latents.shape[0] , _a , _a )
for i, t in enumerate(self.progress_bar(_a ) ):
# expand the latents if we are doing classifier free guidance
_A : List[str] = torch.cat([latents] * 2 ) if do_classifier_free_guidance else latents
_A : int = self.scheduler.scale_model_input(_a , _a )
_A : Tuple = self.prior(
_a , timestep=_a , proj_embedding=_a , ).predicted_image_embedding
# remove the variance
_A , _A : Optional[Any] = noise_pred.split(
scaled_model_input.shape[2] , dim=2 ) # batch_size, num_embeddings, embedding_dim
if do_classifier_free_guidance is not None:
_A , _A : Dict = noise_pred.chunk(2 )
_A : Tuple = noise_pred_uncond + guidance_scale * (noise_pred - noise_pred_uncond)
_A : int = self.scheduler.step(
_a , timestep=_a , sample=_a , ).prev_sample
if output_type == "latent":
return ShapEPipelineOutput(images=_a )
_A : List[str] = []
for i, latent in enumerate(_a ):
print()
_A : List[str] = self.renderer.decode(
latent[None, :] , _a , size=_a , ray_batch_size=4096 , n_coarse_samples=64 , n_fine_samples=128 , )
images.append(_a )
_A : List[Any] = torch.stack(_a )
if output_type not in ["np", "pil"]:
raise ValueError(F'''Only the output types `pil` and `np` are supported not output_type={output_type}''' )
_A : List[str] = images.cpu().numpy()
if output_type == "pil":
_A : List[Any] = [self.numpy_to_pil(_a ) for image in images]
# Offload last model to CPU
if hasattr(self , """final_offload_hook""" ) and self.final_offload_hook is not None:
self.final_offload_hook.offload()
if not return_dict:
return (images,)
return ShapEPipelineOutput(images=_a )
| 26 | 1 |
import inspect
import tempfile
import unittest
from huggingface_hub import hf_hub_download
from transformers import is_torch_available
from transformers.testing_utils import is_flaky, require_torch, slow, torch_device
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
_snake_case = 1e-4
if is_torch_available():
import torch
from transformers import AutoformerConfig, AutoformerForPrediction, AutoformerModel
from transformers.models.autoformer.modeling_autoformer import AutoformerDecoder, AutoformerEncoder
@require_torch
class lowercase :
def __init__( self , _a , _a=16 , _a=13 , _a=7 , _a=14 , _a=10 , _a=19 , _a=5 , _a=4 , _a=True , _a=16 , _a=2 , _a=4 , _a=4 , _a="gelu" , _a=0.1 , _a=0.1 , _a=[1, 2, 3, 4, 5] , _a=25 , _a=5 , ) -> Optional[Any]:
_A : str = d_model
_A : Any = parent
_A : List[str] = batch_size
_A : Any = prediction_length
_A : str = context_length
_A : Any = cardinality
_A : str = num_time_features
_A : str = lags_sequence
_A : List[Any] = embedding_dimension
_A : int = is_training
_A : Tuple = hidden_size
_A : Any = num_hidden_layers
_A : Optional[Any] = num_attention_heads
_A : Tuple = intermediate_size
_A : List[Any] = hidden_act
_A : Tuple = hidden_dropout_prob
_A : Optional[Any] = attention_probs_dropout_prob
_A : Any = context_length
_A : str = prediction_length + label_length
_A : int = label_length
_A : List[str] = moving_average
_A : Dict = autocorrelation_factor
def a__ ( self ) -> List[str]:
return AutoformerConfig(
d_model=self.d_model , encoder_layers=self.num_hidden_layers , decoder_layers=self.num_hidden_layers , encoder_attention_heads=self.num_attention_heads , decoder_attention_heads=self.num_attention_heads , encoder_ffn_dim=self.intermediate_size , decoder_ffn_dim=self.intermediate_size , dropout=self.hidden_dropout_prob , attention_dropout=self.attention_probs_dropout_prob , prediction_length=self.prediction_length , context_length=self.context_length , label_length=self.label_length , lags_sequence=self.lags_sequence , num_time_features=self.num_time_features , num_static_categorical_features=1 , cardinality=[self.cardinality] , embedding_dimension=[self.embedding_dimension] , moving_average=self.moving_average , )
def a__ ( self , _a ) -> Optional[int]:
_A : int = config.context_length + max(config.lags_sequence )
_A : Optional[int] = ids_tensor([self.batch_size, 1] , config.cardinality[0] )
_A : List[str] = floats_tensor([self.batch_size, _past_length, config.num_time_features] )
_A : Optional[Any] = floats_tensor([self.batch_size, _past_length] )
_A : Optional[Any] = floats_tensor([self.batch_size, _past_length] ) > 0.5
# decoder inputs
_A : Union[str, Any] = floats_tensor([self.batch_size, config.prediction_length, config.num_time_features] )
_A : Union[str, Any] = floats_tensor([self.batch_size, config.prediction_length] )
_A : str = {
"""past_values""": past_values,
"""static_categorical_features""": static_categorical_features,
"""past_time_features""": past_time_features,
"""past_observed_mask""": past_observed_mask,
"""future_time_features""": future_time_features,
"""future_values""": future_values,
}
return inputs_dict
def a__ ( self ) -> Tuple:
_A : List[Any] = self.get_config()
_A : int = self.prepare_autoformer_inputs_dict(_a )
return config, inputs_dict
def a__ ( self ) -> Optional[int]:
_A , _A : Tuple = self.prepare_config_and_inputs()
return config, inputs_dict
def a__ ( self , _a , _a ) -> Optional[Any]:
_A : Dict = AutoformerModel(config=_a ).to(_a ).eval()
_A : int = model(**_a )
_A : str = outputs.encoder_last_hidden_state
_A : Optional[Any] = outputs.last_hidden_state
with tempfile.TemporaryDirectory() as tmpdirname:
_A : str = model.get_encoder()
encoder.save_pretrained(_a )
_A : Optional[Any] = AutoformerEncoder.from_pretrained(_a ).to(_a )
_A , _A , _A , _A , _A : Optional[int] = model.create_network_inputs(**_a )
_A , _A : str = model.decomposition_layer(transformer_inputs[:, : config.context_length, ...] )
_A : Union[str, Any] = torch.cat(
(transformer_inputs[:, : config.context_length, ...], feature[:, : config.context_length, ...]) , dim=-1 , )
_A : str = encoder(inputs_embeds=_a )[0]
self.parent.assertTrue((encoder_last_hidden_state_a - encoder_last_hidden_state).abs().max().item() < 1e-3 )
_A : str = (
torch.mean(transformer_inputs[:, : config.context_length, ...] , dim=1 )
.unsqueeze(1 )
.repeat(1 , config.prediction_length , 1 )
)
_A : Optional[int] = torch.zeros(
[transformer_inputs.shape[0], config.prediction_length, transformer_inputs.shape[2]] , device=enc_input.device , )
_A : Tuple = torch.cat(
(
torch.cat((seasonal_input[:, -config.label_length :, ...], zeros) , dim=1 ),
feature[:, config.context_length - config.label_length :, ...],
) , dim=-1 , )
_A : int = torch.cat(
(
torch.cat((trend_input[:, -config.label_length :, ...], mean) , dim=1 ),
feature[:, config.context_length - config.label_length :, ...],
) , dim=-1 , )
with tempfile.TemporaryDirectory() as tmpdirname:
_A : Tuple = model.get_decoder()
decoder.save_pretrained(_a )
_A : Tuple = AutoformerDecoder.from_pretrained(_a ).to(_a )
_A : List[Any] = decoder(
trend=_a , inputs_embeds=_a , encoder_hidden_states=_a , )[0]
self.parent.assertTrue((last_hidden_state_a - last_hidden_state).abs().max().item() < 1e-3 )
@require_torch
class lowercase ( UpperCamelCase__,UpperCamelCase__,unittest.TestCase ):
_a = (AutoformerModel, AutoformerForPrediction) if is_torch_available() else ()
_a = (AutoformerForPrediction,) if is_torch_available() else ()
_a = {"feature-extraction": AutoformerModel} if is_torch_available() else {}
_a = False
_a = False
_a = False
_a = False
_a = False
_a = False
def a__ ( self ) -> Dict:
_A : Optional[int] = AutoformerModelTester(self )
_A : Union[str, Any] = ConfigTester(self , config_class=_a , has_text_modality=_a )
def a__ ( self ) -> int:
self.config_tester.run_common_tests()
def a__ ( self ) -> Optional[int]:
_A , _A : Any = self.model_tester.prepare_config_and_inputs()
for model_class in self.all_model_classes:
_A : Dict = model_class(_a )
with tempfile.TemporaryDirectory() as tmpdirname:
model.save_pretrained(_a )
_A , _A : Any = model_class.from_pretrained(_a , output_loading_info=_a )
self.assertEqual(info["""missing_keys"""] , [] )
def a__ ( self ) -> str:
_A : Dict = self.model_tester.prepare_config_and_inputs_for_common()
self.model_tester.check_encoder_decoder_model_standalone(*_a )
@unittest.skip(reason="""Model has no tokens embeddings""" )
def a__ ( self ) -> Optional[int]:
pass
def a__ ( self ) -> str:
_A : Union[str, Any] = inspect.signature(getattr(_a , """forward""" ) )
# The main input is the name of the argument after `self`
_A : Union[str, Any] = list(model_signature.parameters.keys() )[1]
self.assertEqual(AutoformerModel.main_input_name , _a )
def a__ ( self ) -> List[Any]:
_A , _A : str = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
_A : str = model_class(_a )
_A : List[Any] = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
_A : List[str] = [*signature.parameters.keys()]
_A : Tuple = [
"""past_values""",
"""past_time_features""",
"""past_observed_mask""",
"""static_categorical_features""",
"""static_real_features""",
"""future_values""",
"""future_time_features""",
]
if model.__class__.__name__ in ["AutoformerForPrediction"]:
expected_arg_names.append("""future_observed_mask""" )
expected_arg_names.extend(
[
"""decoder_attention_mask""",
"""head_mask""",
"""decoder_head_mask""",
"""cross_attn_head_mask""",
"""encoder_outputs""",
"""past_key_values""",
"""output_hidden_states""",
"""output_attentions""",
"""use_cache""",
"""return_dict""",
] )
self.assertListEqual(arg_names[: len(_a )] , _a )
def a__ ( self ) -> Optional[Any]:
_A , _A : Optional[int] = self.model_tester.prepare_config_and_inputs_for_common()
_A : Any = True
_A : str = getattr(self.model_tester , """seq_length""" , _a )
_A : Dict = getattr(self.model_tester , """decoder_seq_length""" , _a )
_A : str = getattr(self.model_tester , """encoder_seq_length""" , _a )
_A : List[Any] = getattr(self.model_tester , """d_model""" , _a )
_A : Optional[int] = getattr(self.model_tester , """num_attention_heads""" , _a )
_A : List[str] = d_model // num_attention_heads
for model_class in self.all_model_classes:
_A : Optional[Any] = True
_A : List[str] = False
_A : Optional[int] = True
_A : Union[str, Any] = model_class(_a )
model.to(_a )
model.eval()
with torch.no_grad():
_A : List[Any] = model(**self._prepare_for_class(_a , _a ) )
_A : List[str] = outputs.encoder_attentions if config.is_encoder_decoder else outputs.attentions
self.assertEqual(len(_a ) , self.model_tester.num_hidden_layers )
# check that output_attentions also work using config
del inputs_dict["output_attentions"]
_A : Dict = True
_A : List[Any] = model_class(_a )
model.to(_a )
model.eval()
with torch.no_grad():
_A : int = model(**self._prepare_for_class(_a , _a ) )
_A : Tuple = outputs.encoder_attentions
self.assertEqual(len(_a ) , self.model_tester.num_hidden_layers )
self.assertListEqual(
list(attentions[0].shape[-3:] ) , [self.model_tester.num_attention_heads, encoder_seq_length, dim] , )
_A : List[str] = len(_a )
_A : int = 7
if "last_hidden_state" in outputs:
correct_outlen += 1
if "trend" in outputs:
correct_outlen += 1
if "past_key_values" in outputs:
correct_outlen += 1 # past_key_values have been returned
if "loss" in outputs:
correct_outlen += 1
if "params" in outputs:
correct_outlen += 1
self.assertEqual(_a , _a )
# decoder attentions
_A : Dict = outputs.decoder_attentions
self.assertIsInstance(_a , (list, tuple) )
self.assertEqual(len(_a ) , self.model_tester.num_hidden_layers )
self.assertListEqual(
list(decoder_attentions[0].shape[-3:] ) , [self.model_tester.num_attention_heads, decoder_seq_length, dim] , )
# cross attentions
_A : Optional[Any] = outputs.cross_attentions
self.assertIsInstance(_a , (list, tuple) )
self.assertEqual(len(_a ) , self.model_tester.num_hidden_layers )
self.assertListEqual(
list(cross_attentions[0].shape[-3:] ) , [self.model_tester.num_attention_heads, decoder_seq_length, dim] , )
# Check attention is always last and order is fine
_A : Dict = True
_A : Any = True
_A : str = model_class(_a )
model.to(_a )
model.eval()
with torch.no_grad():
_A : List[str] = model(**self._prepare_for_class(_a , _a ) )
self.assertEqual(out_len + 2 , len(_a ) )
_A : Optional[Any] = outputs.encoder_attentions if config.is_encoder_decoder else outputs.attentions
self.assertEqual(len(_a ) , self.model_tester.num_hidden_layers )
self.assertListEqual(
list(self_attentions[0].shape[-3:] ) , [self.model_tester.num_attention_heads, encoder_seq_length, dim] , )
@is_flaky()
def a__ ( self ) -> int:
super().test_retain_grad_hidden_states_attentions()
def lowerCAmelCase_ ( snake_case_="train-batch.pt" ):
_A : Optional[int] = hf_hub_download(repo_id="""hf-internal-testing/tourism-monthly-batch""",filename=snake_case_,repo_type="""dataset""" )
_A : List[str] = torch.load(snake_case_,map_location=snake_case_ )
return batch
@require_torch
@slow
class lowercase ( unittest.TestCase ):
def a__ ( self ) -> Any:
_A : Optional[int] = AutoformerModel.from_pretrained("""huggingface/autoformer-tourism-monthly""" ).to(_a )
_A : Any = prepare_batch()
with torch.no_grad():
_A : Union[str, Any] = model(
past_values=batch["""past_values"""] , past_time_features=batch["""past_time_features"""] , past_observed_mask=batch["""past_observed_mask"""] , static_categorical_features=batch["""static_categorical_features"""] , future_values=batch["""future_values"""] , future_time_features=batch["""future_time_features"""] , )[0]
_A : List[Any] = torch.Size(
(64, model.config.prediction_length + model.config.label_length, model.config.feature_size) )
self.assertEqual(output.shape , _a )
_A : str = torch.tensor(
[[0.3593, -1.3398, 0.6330], [0.2279, 1.5396, -0.1792], [0.0450, 1.3225, -0.2335]] , device=_a )
self.assertTrue(torch.allclose(output[0, :3, :3] , _a , atol=_a ) )
def a__ ( self ) -> Optional[Any]:
_A : int = AutoformerForPrediction.from_pretrained("""huggingface/autoformer-tourism-monthly""" ).to(_a )
_A : Optional[Any] = prepare_batch("""val-batch.pt""" )
with torch.no_grad():
_A : List[str] = model(
past_values=batch["""past_values"""] , past_time_features=batch["""past_time_features"""] , past_observed_mask=batch["""past_observed_mask"""] , static_categorical_features=batch["""static_categorical_features"""] , ).encoder_last_hidden_state
_A : Optional[int] = torch.Size((64, model.config.context_length, model.config.d_model) )
self.assertEqual(output.shape , _a )
_A : Tuple = torch.tensor(
[[-0.0734, -0.9036, 0.8358], [4.7186, 2.4113, 1.9581], [1.7953, 2.3558, 1.2970]] , device=_a )
self.assertTrue(torch.allclose(output[0, :3, :3] , _a , atol=_a ) )
def a__ ( self ) -> List[str]:
_A : Union[str, Any] = AutoformerForPrediction.from_pretrained("""huggingface/autoformer-tourism-monthly""" ).to(_a )
_A : Optional[int] = prepare_batch("""val-batch.pt""" )
with torch.no_grad():
_A : str = model.generate(
static_categorical_features=batch["""static_categorical_features"""] , past_time_features=batch["""past_time_features"""] , past_values=batch["""past_values"""] , future_time_features=batch["""future_time_features"""] , past_observed_mask=batch["""past_observed_mask"""] , )
_A : str = torch.Size((64, model.config.num_parallel_samples, model.config.prediction_length) )
self.assertEqual(outputs.sequences.shape , _a )
_A : int = torch.tensor([3130.6763, 4056.5293, 7053.0786] , device=_a )
_A : Dict = outputs.sequences.mean(dim=1 )
self.assertTrue(torch.allclose(mean_prediction[0, -3:] , _a , rtol=1e-1 ) )
| 26 |
import argparse
import collections
import json
from pathlib import Path
import requests
import torch
import yaml
from huggingface_hub import hf_hub_download
from PIL import Image
from transformers import (
MobileViTImageProcessor,
MobileViTVaConfig,
MobileViTVaForImageClassification,
MobileViTVaForSemanticSegmentation,
)
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
def lowerCAmelCase_ ( snake_case_ ):
print("""Loading config file...""" )
def flatten_yaml_as_dict(snake_case_,snake_case_="",snake_case_="." ):
_A : Union[str, Any] = []
for k, v in d.items():
_A : Optional[int] = parent_key + sep + k if parent_key else k
if isinstance(snake_case_,collections.abc.MutableMapping ):
items.extend(flatten_yaml_as_dict(snake_case_,snake_case_,sep=snake_case_ ).items() )
else:
items.append((new_key, v) )
return dict(snake_case_ )
_A : List[Any] = argparse.Namespace()
with open(snake_case_,"""r""" ) as yaml_file:
try:
_A : List[Any] = yaml.load(snake_case_,Loader=yaml.FullLoader )
_A : Optional[int] = flatten_yaml_as_dict(snake_case_ )
for k, v in flat_cfg.items():
setattr(snake_case_,snake_case_,snake_case_ )
except yaml.YAMLError as exc:
logger.error("""Error while loading config file: {}. Error message: {}""".format(snake_case_,str(snake_case_ ) ) )
return config
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Optional[Any] = MobileViTVaConfig()
_A : Tuple = False
# dataset
if task_name.startswith("""imagenet1k_""" ):
_A : Dict = 1000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : int = 384
else:
_A : int = 256
_A : List[str] = """imagenet-1k-id2label.json"""
elif task_name.startswith("""imagenet21k_to_1k_""" ):
_A : Union[str, Any] = 21000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : str = 384
else:
_A : List[Any] = 256
_A : List[str] = """imagenet-22k-id2label.json"""
elif task_name.startswith("""ade20k_""" ):
_A : int = 151
_A : int = 512
_A : Optional[int] = """ade20k-id2label.json"""
_A : Any = True
elif task_name.startswith("""voc_""" ):
_A : List[Any] = 21
_A : Dict = 512
_A : Dict = """pascal-voc-id2label.json"""
_A : int = True
# orig_config
_A : Any = load_orig_config_file(snake_case_ )
assert getattr(snake_case_,"""model.classification.name""",-1 ) == "mobilevit_v2", "Invalid model"
_A : List[Any] = getattr(snake_case_,"""model.classification.mitv2.width_multiplier""",1.0 )
assert (
getattr(snake_case_,"""model.classification.mitv2.attn_norm_layer""",-1 ) == "layer_norm_2d"
), "Norm layers other than layer_norm_2d is not supported"
_A : str = getattr(snake_case_,"""model.classification.activation.name""","""swish""" )
# config.image_size == getattr(orig_config, 'sampler.bs.crop_size_width', 256)
if is_segmentation_model:
_A : Optional[int] = getattr(snake_case_,"""model.segmentation.output_stride""",16 )
if "_deeplabv3" in task_name:
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_rates""",[12, 24, 36] )
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_out_channels""",512 )
_A : str = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_dropout""",0.1 )
# id2label
_A : List[Any] = """huggingface/label-files"""
_A : List[Any] = json.load(open(hf_hub_download(snake_case_,snake_case_,repo_type="""dataset""" ),"""r""" ) )
_A : str = {int(snake_case_ ): v for k, v in idalabel.items()}
_A : str = idalabel
_A : Dict = {v: k for k, v in idalabel.items()}
return config
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Any = dct.pop(snake_case_ )
_A : Union[str, Any] = val
def lowerCAmelCase_ ( snake_case_,snake_case_=False ):
if base_model:
_A : Optional[int] = """"""
else:
_A : Dict = """mobilevitv2."""
_A : int = []
for k in state_dict.keys():
if k[:8] == "encoder.":
_A : Any = k[8:]
else:
_A : List[str] = k
if ".block." in k:
_A : Any = k_new.replace(""".block.""",""".""" )
if ".conv." in k:
_A : List[Any] = k_new.replace(""".conv.""",""".convolution.""" )
if ".norm." in k:
_A : Any = k_new.replace(""".norm.""",""".normalization.""" )
if "conv_1." in k:
_A : int = k_new.replace("""conv_1.""",f'''{model_prefix}conv_stem.''' )
for i in [1, 2]:
if f'''layer_{i}.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.''',f'''{model_prefix}encoder.layer.{i-1}.layer.''' )
if ".exp_1x1." in k:
_A : Tuple = k_new.replace(""".exp_1x1.""",""".expand_1x1.""" )
if ".red_1x1." in k:
_A : Optional[int] = k_new.replace(""".red_1x1.""",""".reduce_1x1.""" )
for i in [3, 4, 5]:
if f'''layer_{i}.0.''' in k:
_A : Optional[int] = k_new.replace(f'''layer_{i}.0.''',f'''{model_prefix}encoder.layer.{i-1}.downsampling_layer.''' )
if f'''layer_{i}.1.local_rep.0.''' in k:
_A : Union[str, Any] = k_new.replace(f'''layer_{i}.1.local_rep.0.''',f'''{model_prefix}encoder.layer.{i-1}.conv_kxk.''' )
if f'''layer_{i}.1.local_rep.1.''' in k:
_A : str = k_new.replace(f'''layer_{i}.1.local_rep.1.''',f'''{model_prefix}encoder.layer.{i-1}.conv_1x1.''' )
for i in [3, 4, 5]:
if i == 3:
_A : Optional[int] = [0, 1]
elif i == 4:
_A : Union[str, Any] = [0, 1, 2, 3]
elif i == 5:
_A : Optional[Any] = [0, 1, 2]
for j in j_in:
if f'''layer_{i}.1.global_rep.{j}.''' in k:
_A : Union[str, Any] = k_new.replace(
f'''layer_{i}.1.global_rep.{j}.''',f'''{model_prefix}encoder.layer.{i-1}.transformer.layer.{j}.''' )
if f'''layer_{i}.1.global_rep.{j+1}.''' in k:
_A : List[str] = k_new.replace(
f'''layer_{i}.1.global_rep.{j+1}.''',f'''{model_prefix}encoder.layer.{i-1}.layernorm.''' )
if f'''layer_{i}.1.conv_proj.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.1.conv_proj.''',f'''{model_prefix}encoder.layer.{i-1}.conv_projection.''' )
if "pre_norm_attn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_attn.0.""","""layernorm_before.""" )
if "pre_norm_attn.1." in k:
_A : str = k_new.replace("""pre_norm_attn.1.""","""attention.""" )
if "pre_norm_ffn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_ffn.0.""","""layernorm_after.""" )
if "pre_norm_ffn.1." in k:
_A : Dict = k_new.replace("""pre_norm_ffn.1.""","""ffn.conv1.""" )
if "pre_norm_ffn.3." in k:
_A : List[str] = k_new.replace("""pre_norm_ffn.3.""","""ffn.conv2.""" )
if "classifier.1." in k:
_A : List[str] = k_new.replace("""classifier.1.""","""classifier.""" )
if "seg_head." in k:
_A : List[Any] = k_new.replace("""seg_head.""","""segmentation_head.""" )
if ".aspp_layer." in k:
_A : List[Any] = k_new.replace(""".aspp_layer.""",""".""" )
if ".aspp_pool." in k:
_A : Optional[Any] = k_new.replace(""".aspp_pool.""",""".""" )
rename_keys.append((k, k_new) )
return rename_keys
def lowerCAmelCase_ ( snake_case_ ):
_A : Tuple = []
for k in state_dict.keys():
if k.startswith("""seg_head.aux_head.""" ):
keys_to_ignore.append(snake_case_ )
for k in keys_to_ignore:
state_dict.pop(snake_case_,snake_case_ )
def lowerCAmelCase_ ( ):
_A : Dict = """http://images.cocodataset.org/val2017/000000039769.jpg"""
# url = "https://cdn.britannica.com/86/141086-050-9D7C75EE/Gulfstream-G450-business-jet-passengers.jpg"
_A : List[Any] = Image.open(requests.get(snake_case_,stream=snake_case_ ).raw )
return im
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : List[Any] = get_mobilevitva_config(snake_case_,snake_case_ )
# load original state_dict
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
# load huggingface model
if task_name.startswith("""ade20k_""" ) or task_name.startswith("""voc_""" ):
_A : Optional[Any] = MobileViTVaForSemanticSegmentation(snake_case_ ).eval()
_A : str = False
else:
_A : int = MobileViTVaForImageClassification(snake_case_ ).eval()
_A : List[Any] = False
# remove and rename some keys of load the original model
_A : List[Any] = checkpoint
remove_unused_keys(snake_case_ )
_A : Optional[Any] = create_rename_keys(snake_case_,base_model=snake_case_ )
for rename_key_src, rename_key_dest in rename_keys:
rename_key(snake_case_,snake_case_,snake_case_ )
# load modified state_dict
model.load_state_dict(snake_case_ )
# Check outputs on an image, prepared by MobileViTImageProcessor
_A : str = MobileViTImageProcessor(crop_size=config.image_size,size=config.image_size + 32 )
_A : List[Any] = image_processor(images=prepare_img(),return_tensors="""pt""" )
_A : Optional[Any] = model(**snake_case_ )
# verify classification model
if task_name.startswith("""imagenet""" ):
_A : List[Any] = outputs.logits
_A : Optional[int] = logits.argmax(-1 ).item()
print("""Predicted class:""",model.config.idalabel[predicted_class_idx] )
if task_name.startswith("""imagenet1k_256""" ) and config.width_multiplier == 1.0:
# expected_logits for base variant
_A : int = torch.tensor([-1.63_36e00, -7.32_04e-02, -5.18_83e-01] )
assert torch.allclose(logits[0, :3],snake_case_,atol=1e-4 )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ )
print(f'''Saving model {task_name} to {pytorch_dump_folder_path}''' )
model.save_pretrained(snake_case_ )
print(f'''Saving image processor to {pytorch_dump_folder_path}''' )
image_processor.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
"--task",
default="imagenet1k_256",
type=str,
help=(
"Name of the task for which the MobileViTV2 model you'd like to convert is trained on . "
"\n Classification (ImageNet-1k)\n - MobileViTV2 (256x256) : imagenet1k_256\n - MobileViTV2 (Trained on 256x256 and Finetuned on 384x384) : imagenet1k_384\n - MobileViTV2 (Trained on ImageNet-21k and Finetuned on ImageNet-1k 256x256) :\n imagenet21k_to_1k_256\n - MobileViTV2 (Trained on ImageNet-21k, Finetuned on ImageNet-1k 256x256, and Finetuned on\n ImageNet-1k 384x384) : imagenet21k_to_1k_384\n Segmentation\n - ADE20K Dataset : ade20k_deeplabv3\n - Pascal VOC 2012 Dataset: voc_deeplabv3\n "
),
choices=[
"imagenet1k_256",
"imagenet1k_384",
"imagenet21k_to_1k_256",
"imagenet21k_to_1k_384",
"ade20k_deeplabv3",
"voc_deeplabv3",
],
)
parser.add_argument(
"--orig_checkpoint_path", required=True, type=str, help="Path to the original state dict (.pt file)."
)
parser.add_argument("--orig_config_path", required=True, type=str, help="Path to the original config file.")
parser.add_argument(
"--pytorch_dump_folder_path", required=True, type=str, help="Path to the output PyTorch model directory."
)
_snake_case = parser.parse_args()
convert_mobilevitva_checkpoint(
args.task, args.orig_checkpoint_path, args.orig_config_path, args.pytorch_dump_folder_path
)
| 26 | 1 |
from math import atan, cos, radians, sin, tan
from .haversine_distance import haversine_distance
_snake_case = 6_3_7_8_1_3_7.0
_snake_case = 6_3_5_6_7_5_2.3_1_4_2_4_5
_snake_case = 6378137
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : str = (AXIS_A - AXIS_B) / AXIS_A
# Parametric latitudes
# https://en.wikipedia.org/wiki/Latitude#Parametric_(or_reduced)_latitude
_A : Union[str, Any] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : Optional[int] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
# Compute central angle between two points
# using haversine theta. sigma = haversine_distance / equatorial radius
_A : List[Any] = haversine_distance(snake_case_,snake_case_,snake_case_,snake_case_ ) / EQUATORIAL_RADIUS
# Intermediate P and Q values
_A : Dict = (b_lata + b_lata) / 2
_A : int = (b_lata - b_lata) / 2
# Intermediate X value
# X = (sigma - sin(sigma)) * sin^2Pcos^2Q / cos^2(sigma/2)
_A : Union[str, Any] = (sin(snake_case_ ) ** 2) * (cos(snake_case_ ) ** 2)
_A : str = cos(sigma / 2 ) ** 2
_A : List[str] = (sigma - sin(snake_case_ )) * (x_numerator / x_demonimator)
# Intermediate Y value
# Y = (sigma + sin(sigma)) * cos^2Psin^2Q / sin^2(sigma/2)
_A : List[str] = (cos(snake_case_ ) ** 2) * (sin(snake_case_ ) ** 2)
_A : Optional[Any] = sin(sigma / 2 ) ** 2
_A : int = (sigma + sin(snake_case_ )) * (y_numerator / y_denominator)
return EQUATORIAL_RADIUS * (sigma - ((flattening / 2) * (x_value + y_value)))
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 |
import torch
from diffusers import DPMSolverSDEScheduler
from diffusers.utils import torch_device
from diffusers.utils.testing_utils import require_torchsde
from .test_schedulers import SchedulerCommonTest
@require_torchsde
class lowercase ( UpperCamelCase__ ):
_a = (DPMSolverSDEScheduler,)
_a = 1_0
def a__ ( self , **_a ) -> Optional[Any]:
_A : str = {
"""num_train_timesteps""": 1100,
"""beta_start""": 0.0001,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
"""noise_sampler_seed""": 0,
}
config.update(**_a )
return config
def a__ ( self ) -> Tuple:
for timesteps in [10, 50, 100, 1000]:
self.check_over_configs(num_train_timesteps=_a )
def a__ ( self ) -> Optional[int]:
for beta_start, beta_end in zip([0.00001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ):
self.check_over_configs(beta_start=_a , beta_end=_a )
def a__ ( self ) -> Any:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=_a )
def a__ ( self ) -> Optional[int]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=_a )
def a__ ( self ) -> Optional[int]:
_A : Any = self.scheduler_classes[0]
_A : List[str] = self.get_scheduler_config()
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Dict = self.dummy_model()
_A : Any = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Dict = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : str = model(_a , _a )
_A : List[Any] = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Dict = torch.sum(torch.abs(_a ) )
_A : Dict = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.47821044921875 ) < 1e-2
assert abs(result_mean.item() - 0.2178705964565277 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59352111816406 ) < 1e-2
assert abs(result_mean.item() - 0.22342906892299652 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Optional[Any]:
_A : Dict = self.scheduler_classes[0]
_A : Optional[int] = self.get_scheduler_config(prediction_type="""v_prediction""" )
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Tuple = self.dummy_model()
_A : int = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Tuple = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : int = scheduler.scale_model_input(_a , _a )
_A : Tuple = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Optional[Any] = torch.sum(torch.abs(_a ) )
_A : List[Any] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 124.77149200439453 ) < 1e-2
assert abs(result_mean.item() - 0.16226289014816284 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 128.1663360595703 ) < 1e-2
assert abs(result_mean.item() - 0.16688326001167297 ) < 1e-3
else:
assert abs(result_sum.item() - 119.8487548828125 ) < 1e-2
assert abs(result_mean.item() - 0.1560530662536621 ) < 1e-3
def a__ ( self ) -> List[str]:
_A : Union[str, Any] = self.scheduler_classes[0]
_A : List[Any] = self.get_scheduler_config()
_A : List[str] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Union[str, Any] = self.dummy_model()
_A : Optional[Any] = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
_A : int = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Dict = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : str = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.46957397460938 ) < 1e-2
assert abs(result_mean.item() - 0.21805934607982635 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59353637695312 ) < 1e-2
assert abs(result_mean.item() - 0.22342908382415771 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Union[str, Any]:
_A : List[Any] = self.scheduler_classes[0]
_A : Optional[Any] = self.get_scheduler_config()
_A : int = scheduler_class(**_a , use_karras_sigmas=_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Optional[Any] = self.dummy_model()
_A : Dict = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
_A : str = sample.to(_a )
for t in scheduler.timesteps:
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : List[str] = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : List[str] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 176.66974135742188 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 177.63653564453125 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
else:
assert abs(result_sum.item() - 170.3135223388672 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
| 26 | 1 |
import unittest
from huggingface_hub import hf_hub_download
from transformers import MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING, VideoMAEFeatureExtractor
from transformers.pipelines import VideoClassificationPipeline, pipeline
from transformers.testing_utils import (
is_pipeline_test,
nested_simplify,
require_decord,
require_tf,
require_torch,
require_torch_or_tf,
require_vision,
)
from .test_pipelines_common import ANY
@is_pipeline_test
@require_torch_or_tf
@require_vision
@require_decord
class lowercase ( unittest.TestCase ):
_a = MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING
def a__ ( self , _a , _a , _a ) -> Any:
_A : List[Any] = hf_hub_download(
repo_id="""nateraw/video-demo""" , filename="""archery.mp4""" , repo_type="""dataset""" )
_A : Tuple = VideoClassificationPipeline(model=_a , image_processor=_a , top_k=2 )
_A : Optional[Any] = [
example_video_filepath,
"""https://huggingface.co/datasets/nateraw/video-demo/resolve/main/archery.mp4""",
]
return video_classifier, examples
def a__ ( self , _a , _a ) -> Any:
for example in examples:
_A : Optional[int] = video_classifier(_a )
self.assertEqual(
_a , [
{"""score""": ANY(_a ), """label""": ANY(_a )},
{"""score""": ANY(_a ), """label""": ANY(_a )},
] , )
@require_torch
def a__ ( self ) -> str:
_A : Tuple = """hf-internal-testing/tiny-random-VideoMAEForVideoClassification"""
_A : Tuple = VideoMAEFeatureExtractor(
size={"""shortest_edge""": 10} , crop_size={"""height""": 10, """width""": 10} )
_A : Any = pipeline(
"""video-classification""" , model=_a , feature_extractor=_a , frame_sampling_rate=4 )
_A : Optional[int] = hf_hub_download(repo_id="""nateraw/video-demo""" , filename="""archery.mp4""" , repo_type="""dataset""" )
_A : Any = video_classifier(_a , top_k=2 )
self.assertEqual(
nested_simplify(_a , decimals=4 ) , [{"""score""": 0.5199, """label""": """LABEL_0"""}, {"""score""": 0.4801, """label""": """LABEL_1"""}] , )
_A : Optional[int] = video_classifier(
[
video_file_path,
video_file_path,
] , top_k=2 , )
self.assertEqual(
nested_simplify(_a , decimals=4 ) , [
[{"""score""": 0.5199, """label""": """LABEL_0"""}, {"""score""": 0.4801, """label""": """LABEL_1"""}],
[{"""score""": 0.5199, """label""": """LABEL_0"""}, {"""score""": 0.4801, """label""": """LABEL_1"""}],
] , )
@require_tf
def a__ ( self ) -> str:
pass
| 26 |
# DISCLAIMER: This file is strongly influenced by https://github.com/yang-song/score_sde_pytorch
import math
from typing import Union
import torch
from ..configuration_utils import ConfigMixin, register_to_config
from ..utils import randn_tensor
from .scheduling_utils import SchedulerMixin
class lowercase ( UpperCamelCase__,UpperCamelCase__ ):
_a = 1
@register_to_config
def __init__( self , _a=2000 , _a=0.1 , _a=20 , _a=1e-3 ) -> List[Any]:
_A : Dict = None
_A : List[Any] = None
_A : Dict = None
def a__ ( self , _a , _a = None ) -> Union[str, Any]:
_A : Union[str, Any] = torch.linspace(1 , self.config.sampling_eps , _a , device=_a )
def a__ ( self , _a , _a , _a , _a=None ) -> Dict:
if self.timesteps is None:
raise ValueError(
"""`self.timesteps` is not set, you need to run 'set_timesteps' after creating the scheduler""" )
# TODO(Patrick) better comments + non-PyTorch
# postprocess model score
_A : Any = (
-0.25 * t**2 * (self.config.beta_max - self.config.beta_min) - 0.5 * t * self.config.beta_min
)
_A : List[Any] = torch.sqrt(1.0 - torch.exp(2.0 * log_mean_coeff ) )
_A : List[str] = std.flatten()
while len(std.shape ) < len(score.shape ):
_A : List[Any] = std.unsqueeze(-1 )
_A : int = -score / std
# compute
_A : Tuple = -1.0 / len(self.timesteps )
_A : str = self.config.beta_min + t * (self.config.beta_max - self.config.beta_min)
_A : List[str] = beta_t.flatten()
while len(beta_t.shape ) < len(x.shape ):
_A : Union[str, Any] = beta_t.unsqueeze(-1 )
_A : Tuple = -0.5 * beta_t * x
_A : Tuple = torch.sqrt(_a )
_A : Dict = drift - diffusion**2 * score
_A : Dict = x + drift * dt
# add noise
_A : Any = randn_tensor(x.shape , layout=x.layout , generator=_a , device=x.device , dtype=x.dtype )
_A : str = x_mean + diffusion * math.sqrt(-dt ) * noise
return x, x_mean
def __len__( self ) -> Optional[Any]:
return self.config.num_train_timesteps
| 26 | 1 |
import argparse
import torch
from torch import nn
from transformers import MBartConfig, MBartForConditionalGeneration
def lowerCAmelCase_ ( snake_case_ ):
_A : List[str] = [
"""encoder.version""",
"""decoder.version""",
"""model.encoder.version""",
"""model.decoder.version""",
"""_float_tensor""",
"""decoder.output_projection.weight""",
]
for k in ignore_keys:
state_dict.pop(snake_case_,snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
_A , _A : List[str] = emb.weight.shape
_A : Any = nn.Linear(snake_case_,snake_case_,bias=snake_case_ )
_A : str = emb.weight.data
return lin_layer
def lowerCAmelCase_ ( snake_case_,snake_case_="facebook/mbart-large-en-ro",snake_case_=False,snake_case_=False ):
_A : Optional[Any] = torch.load(snake_case_,map_location="""cpu""" )["""model"""]
remove_ignore_keys_(snake_case_ )
_A : Optional[int] = state_dict["""encoder.embed_tokens.weight"""].shape[0]
_A : Optional[Any] = MBartConfig.from_pretrained(snake_case_,vocab_size=snake_case_ )
if mbart_aa and finetuned:
_A : Union[str, Any] = """relu"""
_A : str = state_dict["""decoder.embed_tokens.weight"""]
_A : Optional[Any] = MBartForConditionalGeneration(snake_case_ )
model.model.load_state_dict(snake_case_ )
if finetuned:
_A : Dict = make_linear_from_emb(model.model.shared )
return model
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
"fairseq_path", type=str, help="bart.large, bart.large.cnn or a path to a model.pt on local filesystem."
)
parser.add_argument("pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model.")
parser.add_argument(
"--hf_config",
default="facebook/mbart-large-cc25",
type=str,
help="Which huggingface architecture to use: mbart-large",
)
parser.add_argument("--mbart_50", action="store_true", help="whether the model is mMART-50 checkpoint")
parser.add_argument("--finetuned", action="store_true", help="whether the model is a fine-tuned checkpoint")
_snake_case = parser.parse_args()
_snake_case = convert_fairseq_mbart_checkpoint_from_disk(
args.fairseq_path, hf_config_path=args.hf_config, finetuned=args.finetuned, mbart_aa=args.mbart_aa
)
model.save_pretrained(args.pytorch_dump_folder_path)
| 26 |
import os
from shutil import copyfile
from typing import List, Optional, Tuple
from ...tokenization_utils import AddedToken
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import is_sentencepiece_available, logging
if is_sentencepiece_available():
from .tokenization_fnet import FNetTokenizer
else:
_snake_case = None
_snake_case = logging.get_logger(__name__)
_snake_case = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.json"}
_snake_case = {
"vocab_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/spiece.model",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/spiece.model",
},
"tokenizer_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/tokenizer.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/tokenizer.json",
},
}
_snake_case = {
"google/fnet-base": 512,
"google/fnet-large": 512,
}
_snake_case = "▁"
class lowercase ( UpperCamelCase__ ):
_a = VOCAB_FILES_NAMES
_a = PRETRAINED_VOCAB_FILES_MAP
_a = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
_a = ["input_ids", "token_type_ids"]
_a = FNetTokenizer
def __init__( self , _a=None , _a=None , _a=False , _a=True , _a=True , _a="<unk>" , _a="[SEP]" , _a="<pad>" , _a="[CLS]" , _a="[MASK]" , **_a , ) -> Optional[int]:
# Mask token behave like a normal word, i.e. include the space before it and
# is included in the raw text, there should be a match in a non-normalized sentence.
_A : int = (
AddedToken(_a , lstrip=_a , rstrip=_a , normalized=_a )
if isinstance(_a , _a )
else mask_token
)
super().__init__(
_a , tokenizer_file=_a , do_lower_case=_a , remove_space=_a , keep_accents=_a , unk_token=_a , sep_token=_a , pad_token=_a , cls_token=_a , mask_token=_a , **_a , )
_A : Optional[int] = do_lower_case
_A : List[Any] = remove_space
_A : str = keep_accents
_A : int = vocab_file
_A : int = False if not self.vocab_file else True
def a__ ( self , _a , _a = None ) -> List[int]:
_A : str = [self.sep_token_id]
_A : Dict = [self.cls_token_id]
if token_ids_a is None:
return cls + token_ids_a + sep
return cls + token_ids_a + sep + token_ids_a + sep
def a__ ( self , _a , _a = None ) -> List[int]:
_A : Any = [self.sep_token_id]
_A : Any = [self.cls_token_id]
if token_ids_a is None:
return len(cls + token_ids_a + sep ) * [0]
return len(cls + token_ids_a + sep ) * [0] + len(token_ids_a + sep ) * [1]
def a__ ( self , _a , _a = None ) -> Tuple[str]:
if not os.path.isdir(_a ):
logger.error(F'''Vocabulary path ({save_directory}) should be a directory''' )
return
_A : List[str] = os.path.join(
_a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
if os.path.abspath(self.vocab_file ) != os.path.abspath(_a ):
copyfile(self.vocab_file , _a )
return (out_vocab_file,)
| 26 | 1 |
import copy
from ...configuration_utils import PretrainedConfig
from ...utils import logging
from ..bit import BitConfig
_snake_case = logging.get_logger(__name__)
_snake_case = {
"Intel/dpt-large": "https://huggingface.co/Intel/dpt-large/resolve/main/config.json",
# See all DPT models at https://huggingface.co/models?filter=dpt
}
class lowercase ( UpperCamelCase__ ):
_a = "dpt"
def __init__( self , _a=768 , _a=12 , _a=12 , _a=3072 , _a="gelu" , _a=0.0 , _a=0.0 , _a=0.02 , _a=1e-12 , _a=384 , _a=16 , _a=3 , _a=False , _a=True , _a=[2, 5, 8, 11] , _a="project" , _a=[4, 2, 1, 0.5] , _a=[96, 192, 384, 768] , _a=256 , _a=-1 , _a=False , _a=True , _a=0.4 , _a=255 , _a=0.1 , _a=[1, 1024, 24, 24] , _a=[0, 1] , _a=None , **_a , ) -> Optional[Any]:
super().__init__(**_a )
_A : Any = hidden_size
_A : Optional[int] = is_hybrid
if self.is_hybrid:
if backbone_config is None:
logger.info("""Initializing the config with a `BiT` backbone.""" )
_A : Union[str, Any] = {
"""global_padding""": """same""",
"""layer_type""": """bottleneck""",
"""depths""": [3, 4, 9],
"""out_features""": ["""stage1""", """stage2""", """stage3"""],
"""embedding_dynamic_padding""": True,
}
_A : Union[str, Any] = BitConfig(**_a )
elif isinstance(_a , _a ):
logger.info("""Initializing the config with a `BiT` backbone.""" )
_A : Union[str, Any] = BitConfig(**_a )
elif isinstance(_a , _a ):
_A : Optional[Any] = backbone_config
else:
raise ValueError(
F'''backbone_config must be a dictionary or a `PretrainedConfig`, got {backbone_config.__class__}.''' )
_A : Any = backbone_featmap_shape
_A : Optional[Any] = neck_ignore_stages
if readout_type != "project":
raise ValueError("""Readout type must be 'project' when using `DPT-hybrid` mode.""" )
else:
_A : Tuple = None
_A : Union[str, Any] = None
_A : Dict = []
_A : Union[str, Any] = num_hidden_layers
_A : List[str] = num_attention_heads
_A : Union[str, Any] = intermediate_size
_A : Tuple = hidden_act
_A : int = hidden_dropout_prob
_A : Tuple = attention_probs_dropout_prob
_A : List[str] = initializer_range
_A : List[Any] = layer_norm_eps
_A : str = image_size
_A : str = patch_size
_A : List[str] = num_channels
_A : int = qkv_bias
_A : Dict = backbone_out_indices
if readout_type not in ["ignore", "add", "project"]:
raise ValueError("""Readout_type must be one of ['ignore', 'add', 'project']""" )
_A : int = readout_type
_A : List[str] = reassemble_factors
_A : str = neck_hidden_sizes
_A : Union[str, Any] = fusion_hidden_size
_A : List[Any] = head_in_index
_A : List[Any] = use_batch_norm_in_fusion_residual
# auxiliary head attributes (semantic segmentation)
_A : Tuple = use_auxiliary_head
_A : Union[str, Any] = auxiliary_loss_weight
_A : Tuple = semantic_loss_ignore_index
_A : List[Any] = semantic_classifier_dropout
def a__ ( self ) -> Union[str, Any]:
_A : str = copy.deepcopy(self.__dict__ )
if output["backbone_config"] is not None:
_A : Tuple = self.backbone_config.to_dict()
_A : Tuple = self.__class__.model_type
return output
| 26 |
from math import asin, atan, cos, radians, sin, sqrt, tan
_snake_case = 6_3_7_8_1_3_7.0
_snake_case = 6_3_5_6_7_5_2.3_1_4_2_4_5
_snake_case = 6378137
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : Any = (AXIS_A - AXIS_B) / AXIS_A
_A : Optional[int] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : List[str] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : Optional[Any] = radians(snake_case_ )
_A : str = radians(snake_case_ )
# Equation
_A : Dict = sin((phi_a - phi_a) / 2 )
_A : List[str] = sin((lambda_a - lambda_a) / 2 )
# Square both values
sin_sq_phi *= sin_sq_phi
sin_sq_lambda *= sin_sq_lambda
_A : Optional[int] = sqrt(sin_sq_phi + (cos(snake_case_ ) * cos(snake_case_ ) * sin_sq_lambda) )
return 2 * RADIUS * asin(snake_case_ )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
import csv
from collections import defaultdict
from dataclasses import dataclass, field
from typing import List, Optional
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.ticker import ScalarFormatter
from transformers import HfArgumentParser
def lowerCAmelCase_ ( snake_case_=None,snake_case_=None ):
return field(default_factory=lambda: default,metadata=snake_case_ )
@dataclass
class lowercase :
_a = field(
metadata={"help": "The csv file to plot."},)
_a = field(
default=UpperCamelCase__,metadata={"help": "Whether to plot along batch size or sequence length. Defaults to sequence length."},)
_a = field(
default=UpperCamelCase__,metadata={"help": "Whether the csv file has time results or memory results. Defaults to memory results."},)
_a = field(
default=UpperCamelCase__,metadata={"help": "Disable logarithmic scale when plotting"},)
_a = field(
default=UpperCamelCase__,metadata={
"help": "Whether the csv file has training results or inference results. Defaults to inference results."
},)
_a = field(
default=UpperCamelCase__,metadata={"help": "Filename under which the plot will be saved. If unused no plot is saved."},)
_a = list_field(
default=UpperCamelCase__,metadata={"help": "List of model names that are used instead of the ones in the csv file."} )
def lowerCAmelCase_ ( snake_case_ ):
try:
int(snake_case_ )
return True
except ValueError:
return False
def lowerCAmelCase_ ( snake_case_ ):
try:
float(snake_case_ )
return True
except ValueError:
return False
class lowercase :
def __init__( self , _a ) -> Optional[Any]:
_A : int = args
_A : Optional[Any] = defaultdict(lambda: {"bsz": [], "seq_len": [], "result": {}} )
with open(self.args.csv_file , newline="""""" ) as csv_file:
_A : List[str] = csv.DictReader(_a )
for row in reader:
_A : Optional[Any] = row["""model"""]
self.result_dict[model_name]["bsz"].append(int(row["""batch_size"""] ) )
self.result_dict[model_name]["seq_len"].append(int(row["""sequence_length"""] ) )
if can_convert_to_int(row["""result"""] ):
# value is not None
_A : Union[str, Any] = int(row["""result"""] )
elif can_convert_to_float(row["""result"""] ):
# value is not None
_A : int = float(row["""result"""] )
def a__ ( self ) -> Any:
_A , _A : Tuple = plt.subplots()
_A : Union[str, Any] = """Time usage""" if self.args.is_time else """Memory usage"""
_A : List[Any] = title_str + """ for training""" if self.args.is_train else title_str + """ for inference"""
if not self.args.no_log_scale:
# set logarithm scales
ax.set_xscale("""log""" )
ax.set_yscale("""log""" )
for axis in [ax.xaxis, ax.yaxis]:
axis.set_major_formatter(ScalarFormatter() )
for model_name_idx, model_name in enumerate(self.result_dict.keys() ):
_A : Optional[Any] = sorted(set(self.result_dict[model_name]["""bsz"""] ) )
_A : List[Any] = sorted(set(self.result_dict[model_name]["""seq_len"""] ) )
_A : List[str] = self.result_dict[model_name]["""result"""]
((_A) , (_A)) : Dict = (
(batch_sizes, sequence_lengths) if self.args.plot_along_batch else (sequence_lengths, batch_sizes)
)
_A : Union[str, Any] = (
model_name if self.args.short_model_names is None else self.args.short_model_names[model_name_idx]
)
for inner_loop_value in inner_loop_array:
if self.args.plot_along_batch:
_A : int = np.asarray(
[results[(x, inner_loop_value)] for x in x_axis_array if (x, inner_loop_value) in results] , dtype=_a , )
else:
_A : int = np.asarray(
[results[(inner_loop_value, x)] for x in x_axis_array if (inner_loop_value, x) in results] , dtype=np.floataa , )
((_A) , (_A)) : Any = (
("""batch_size""", """len""") if self.args.plot_along_batch else ("""in #tokens""", """bsz""")
)
_A : List[Any] = np.asarray(_a , _a )[: len(_a )]
plt.scatter(
_a , _a , label=F'''{label_model_name} - {inner_loop_label}: {inner_loop_value}''' )
plt.plot(_a , _a , """--""" )
title_str += F''' {label_model_name} vs.'''
_A : Optional[Any] = title_str[:-4]
_A : Optional[int] = """Time in s""" if self.args.is_time else """Memory in MB"""
# plot
plt.title(_a )
plt.xlabel(_a )
plt.ylabel(_a )
plt.legend()
if self.args.figure_png_file is not None:
plt.savefig(self.args.figure_png_file )
else:
plt.show()
def lowerCAmelCase_ ( ):
_A : List[Any] = HfArgumentParser(snake_case_ )
_A : Tuple = parser.parse_args_into_dataclasses()[0]
_A : Dict = Plot(args=snake_case_ )
plot.plot()
if __name__ == "__main__":
main()
| 26 |
from __future__ import absolute_import, division, print_function, unicode_literals
from torch import nn
from torch.nn import CrossEntropyLoss, MSELoss
from transformers import RobertaConfig
from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
from transformers.models.roberta.modeling_roberta import (
ROBERTA_INPUTS_DOCSTRING,
ROBERTA_START_DOCSTRING,
RobertaEmbeddings,
)
from .modeling_highway_bert import BertPreTrainedModel, DeeBertModel, HighwayException, entropy
@add_start_docstrings(
"The RoBERTa Model transformer with early exiting (DeeRoBERTa). ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> Optional[int]:
super().__init__(_a )
_A : Union[str, Any] = RobertaEmbeddings(_a )
self.init_weights()
@add_start_docstrings(
"RoBERTa Model (with early exiting - DeeRoBERTa) with a classifier on top,\n also takes care of multi-layer training. ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> str:
super().__init__(_a )
_A : Any = config.num_labels
_A : Dict = config.num_hidden_layers
_A : List[str] = DeeRobertaModel(_a )
_A : int = nn.Dropout(config.hidden_dropout_prob )
_A : int = nn.Linear(config.hidden_size , self.config.num_labels )
@add_start_docstrings_to_model_forward(_a )
def a__ ( self , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=-1 , _a=False , ) -> Any:
_A : Optional[int] = self.num_layers
try:
_A : List[str] = self.roberta(
_a , attention_mask=_a , token_type_ids=_a , position_ids=_a , head_mask=_a , inputs_embeds=_a , )
_A : List[str] = outputs[1]
_A : List[str] = self.dropout(_a )
_A : Optional[Any] = self.classifier(_a )
_A : List[Any] = (logits,) + outputs[2:] # add hidden states and attention if they are here
except HighwayException as e:
_A : List[Any] = e.message
_A : Optional[int] = e.exit_layer
_A : Optional[int] = outputs[0]
if not self.training:
_A : int = entropy(_a )
_A : int = []
_A : int = []
if labels is not None:
if self.num_labels == 1:
# We are doing regression
_A : Union[str, Any] = MSELoss()
_A : Tuple = loss_fct(logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Dict = loss_fct(logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
# work with highway exits
_A : Optional[Any] = []
for highway_exit in outputs[-1]:
_A : Tuple = highway_exit[0]
if not self.training:
highway_logits_all.append(_a )
highway_entropy.append(highway_exit[2] )
if self.num_labels == 1:
# We are doing regression
_A : List[str] = MSELoss()
_A : Optional[int] = loss_fct(highway_logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Tuple = loss_fct(highway_logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
highway_losses.append(_a )
if train_highway:
_A : Dict = (sum(highway_losses[:-1] ),) + outputs
# exclude the final highway, of course
else:
_A : int = (loss,) + outputs
if not self.training:
_A : Optional[Any] = outputs + ((original_entropy, highway_entropy), exit_layer)
if output_layer >= 0:
_A : Union[str, Any] = (
(outputs[0],) + (highway_logits_all[output_layer],) + outputs[2:]
) # use the highway of the last layer
return outputs # (loss), logits, (hidden_states), (attentions), entropy
| 26 | 1 |
from typing import Any, Callable, Dict, List, Optional, Union
import torch
from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
DiffusionPipeline,
LMSDiscreteScheduler,
PNDMScheduler,
StableDiffusionPipeline,
UNetaDConditionModel,
)
from diffusers.pipelines.stable_diffusion import StableDiffusionPipelineOutput
from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
_snake_case = "CompVis/stable-diffusion-v1-1"
_snake_case = "CompVis/stable-diffusion-v1-2"
_snake_case = "CompVis/stable-diffusion-v1-3"
_snake_case = "CompVis/stable-diffusion-v1-4"
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a , _a , _a , _a , _a , _a = True , ) -> Dict:
super()._init_()
_A : Optional[Any] = StableDiffusionPipeline.from_pretrained(_a )
_A : List[str] = StableDiffusionPipeline.from_pretrained(_a )
_A : Dict = StableDiffusionPipeline.from_pretrained(_a )
_A : Union[str, Any] = StableDiffusionPipeline(
vae=_a , text_encoder=_a , tokenizer=_a , unet=_a , scheduler=_a , safety_checker=_a , feature_extractor=_a , requires_safety_checker=_a , )
self.register_modules(pipelinea=self.pipea , pipelinea=self.pipea , pipelinea=self.pipea , pipelinea=self.pipea )
@property
def a__ ( self ) -> Dict[str, Any]:
return {k: getattr(self , _a ) for k in self.config.keys() if not k.startswith("""_""" )}
def a__ ( self , _a = "auto" ) -> Dict:
if slice_size == "auto":
# half the attention head size is usually a good trade-off between
# speed and memory
_A : int = self.unet.config.attention_head_dim // 2
self.unet.set_attention_slice(_a )
def a__ ( self ) -> Any:
self.enable_attention_slicing(_a )
@torch.no_grad()
def a__ ( self , _a , _a = 512 , _a = 512 , _a = 50 , _a = 7.5 , _a = None , _a = 1 , _a = 0.0 , _a = None , _a = None , _a = "pil" , _a = True , _a = None , _a = 1 , **_a , ) -> List[str]:
return self.pipea(
prompt=_a , height=_a , width=_a , num_inference_steps=_a , guidance_scale=_a , negative_prompt=_a , num_images_per_prompt=_a , eta=_a , generator=_a , latents=_a , output_type=_a , return_dict=_a , callback=_a , callback_steps=_a , **_a , )
@torch.no_grad()
def a__ ( self , _a , _a = 512 , _a = 512 , _a = 50 , _a = 7.5 , _a = None , _a = 1 , _a = 0.0 , _a = None , _a = None , _a = "pil" , _a = True , _a = None , _a = 1 , **_a , ) -> str:
return self.pipea(
prompt=_a , height=_a , width=_a , num_inference_steps=_a , guidance_scale=_a , negative_prompt=_a , num_images_per_prompt=_a , eta=_a , generator=_a , latents=_a , output_type=_a , return_dict=_a , callback=_a , callback_steps=_a , **_a , )
@torch.no_grad()
def a__ ( self , _a , _a = 512 , _a = 512 , _a = 50 , _a = 7.5 , _a = None , _a = 1 , _a = 0.0 , _a = None , _a = None , _a = "pil" , _a = True , _a = None , _a = 1 , **_a , ) -> int:
return self.pipea(
prompt=_a , height=_a , width=_a , num_inference_steps=_a , guidance_scale=_a , negative_prompt=_a , num_images_per_prompt=_a , eta=_a , generator=_a , latents=_a , output_type=_a , return_dict=_a , callback=_a , callback_steps=_a , **_a , )
@torch.no_grad()
def a__ ( self , _a , _a = 512 , _a = 512 , _a = 50 , _a = 7.5 , _a = None , _a = 1 , _a = 0.0 , _a = None , _a = None , _a = "pil" , _a = True , _a = None , _a = 1 , **_a , ) -> List[Any]:
return self.pipea(
prompt=_a , height=_a , width=_a , num_inference_steps=_a , guidance_scale=_a , negative_prompt=_a , num_images_per_prompt=_a , eta=_a , generator=_a , latents=_a , output_type=_a , return_dict=_a , callback=_a , callback_steps=_a , **_a , )
@torch.no_grad()
def a__ ( self , _a , _a = 512 , _a = 512 , _a = 50 , _a = 7.5 , _a = None , _a = 1 , _a = 0.0 , _a = None , _a = None , _a = "pil" , _a = True , _a = None , _a = 1 , **_a , ) -> Union[str, Any]:
_A : List[Any] = """cuda""" if torch.cuda.is_available() else """cpu"""
self.to(_a )
# Checks if the height and width are divisible by 8 or not
if height % 8 != 0 or width % 8 != 0:
raise ValueError(F'''`height` and `width` must be divisible by 8 but are {height} and {width}.''' )
# Get first result from Stable Diffusion Checkpoint v1.1
_A : Dict = self.textaimg_sda_a(
prompt=_a , height=_a , width=_a , num_inference_steps=_a , guidance_scale=_a , negative_prompt=_a , num_images_per_prompt=_a , eta=_a , generator=_a , latents=_a , output_type=_a , return_dict=_a , callback=_a , callback_steps=_a , **_a , )
# Get first result from Stable Diffusion Checkpoint v1.2
_A : Optional[int] = self.textaimg_sda_a(
prompt=_a , height=_a , width=_a , num_inference_steps=_a , guidance_scale=_a , negative_prompt=_a , num_images_per_prompt=_a , eta=_a , generator=_a , latents=_a , output_type=_a , return_dict=_a , callback=_a , callback_steps=_a , **_a , )
# Get first result from Stable Diffusion Checkpoint v1.3
_A : Dict = self.textaimg_sda_a(
prompt=_a , height=_a , width=_a , num_inference_steps=_a , guidance_scale=_a , negative_prompt=_a , num_images_per_prompt=_a , eta=_a , generator=_a , latents=_a , output_type=_a , return_dict=_a , callback=_a , callback_steps=_a , **_a , )
# Get first result from Stable Diffusion Checkpoint v1.4
_A : List[str] = self.textaimg_sda_a(
prompt=_a , height=_a , width=_a , num_inference_steps=_a , guidance_scale=_a , negative_prompt=_a , num_images_per_prompt=_a , eta=_a , generator=_a , latents=_a , output_type=_a , return_dict=_a , callback=_a , callback_steps=_a , **_a , )
# Get all result images into a single list and pass it via StableDiffusionPipelineOutput for final result
return StableDiffusionPipelineOutput([resa[0], resa[0], resa[0], resa[0]] )
| 26 |
from collections import OrderedDict
from typing import Mapping
from ...configuration_utils import PretrainedConfig
from ...onnx import OnnxConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"facebook/xmod-base": "https://huggingface.co/facebook/xmod-base/resolve/main/config.json",
"facebook/xmod-large-prenorm": "https://huggingface.co/facebook/xmod-large-prenorm/resolve/main/config.json",
"facebook/xmod-base-13-125k": "https://huggingface.co/facebook/xmod-base-13-125k/resolve/main/config.json",
"facebook/xmod-base-30-125k": "https://huggingface.co/facebook/xmod-base-30-125k/resolve/main/config.json",
"facebook/xmod-base-30-195k": "https://huggingface.co/facebook/xmod-base-30-195k/resolve/main/config.json",
"facebook/xmod-base-60-125k": "https://huggingface.co/facebook/xmod-base-60-125k/resolve/main/config.json",
"facebook/xmod-base-60-265k": "https://huggingface.co/facebook/xmod-base-60-265k/resolve/main/config.json",
"facebook/xmod-base-75-125k": "https://huggingface.co/facebook/xmod-base-75-125k/resolve/main/config.json",
"facebook/xmod-base-75-269k": "https://huggingface.co/facebook/xmod-base-75-269k/resolve/main/config.json",
}
class lowercase ( UpperCamelCase__ ):
_a = "xmod"
def __init__( self , _a=3_0522 , _a=768 , _a=12 , _a=12 , _a=3072 , _a="gelu" , _a=0.1 , _a=0.1 , _a=512 , _a=2 , _a=0.02 , _a=1e-12 , _a=1 , _a=0 , _a=2 , _a="absolute" , _a=True , _a=None , _a=False , _a=2 , _a=False , _a=True , _a=True , _a=("en_XX",) , _a=None , **_a , ) -> str:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Tuple = vocab_size
_A : Union[str, Any] = hidden_size
_A : Dict = num_hidden_layers
_A : Dict = num_attention_heads
_A : List[Any] = hidden_act
_A : Optional[Any] = intermediate_size
_A : Any = hidden_dropout_prob
_A : str = attention_probs_dropout_prob
_A : Dict = max_position_embeddings
_A : Any = type_vocab_size
_A : List[Any] = initializer_range
_A : int = layer_norm_eps
_A : int = position_embedding_type
_A : Any = use_cache
_A : int = classifier_dropout
_A : int = pre_norm
_A : Optional[Any] = adapter_reduction_factor
_A : List[Any] = adapter_layer_norm
_A : Optional[int] = adapter_reuse_layer_norm
_A : Any = ln_before_adapter
_A : Union[str, Any] = list(_a )
_A : List[Any] = default_language
class lowercase ( UpperCamelCase__ ):
@property
def a__ ( self ) -> Mapping[str, Mapping[int, str]]:
if self.task == "multiple-choice":
_A : Dict = {0: """batch""", 1: """choice""", 2: """sequence"""}
else:
_A : Dict = {0: """batch""", 1: """sequence"""}
return OrderedDict(
[
("""input_ids""", dynamic_axis),
("""attention_mask""", dynamic_axis),
] )
| 26 | 1 |
def lowerCAmelCase_ ( snake_case_ ):
_A : Dict = n ** (1 / 3)
return (val * val * val) == n
if __name__ == "__main__":
print(perfect_cube(27))
print(perfect_cube(4))
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
if n == 0:
return 0
_A : Tuple = float("""-inf""" )
for i in range(1,n + 1 ):
_A : str = max(
snake_case_,prices[i - 1] + naive_cut_rod_recursive(n - i,snake_case_ ) )
return max_revue
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
_A : Dict = [float("""-inf""" ) for _ in range(n + 1 )]
return _top_down_cut_rod_recursive(snake_case_,snake_case_,snake_case_ )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
if max_rev[n] >= 0:
return max_rev[n]
elif n == 0:
return 0
else:
_A : List[str] = float("""-inf""" )
for i in range(1,n + 1 ):
_A : Optional[Any] = max(
snake_case_,prices[i - 1] + _top_down_cut_rod_recursive(n - i,snake_case_,snake_case_ ),)
_A : Tuple = max_revenue
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
# length(max_rev) = n + 1, to accommodate for the revenue obtainable from a rod of
# length 0.
_A : List[Any] = [float("""-inf""" ) for _ in range(n + 1 )]
_A : Any = 0
for i in range(1,n + 1 ):
_A : Optional[Any] = max_rev[i]
for j in range(1,i + 1 ):
_A : int = max(snake_case_,prices[j - 1] + max_rev[i - j] )
_A : int = max_revenue_i
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if n < 0:
_A : Optional[Any] = f'''n must be greater than or equal to 0. Got n = {n}'''
raise ValueError(snake_case_ )
if n > len(snake_case_ ):
_A : Any = (
"""Each integral piece of rod must have a corresponding price. """
f'''Got n = {n} but length of prices = {len(snake_case_ )}'''
)
raise ValueError(snake_case_ )
def lowerCAmelCase_ ( ):
_A : Tuple = [6, 10, 12, 15, 20, 23]
_A : List[Any] = len(snake_case_ )
# the best revenue comes from cutting the rod into 6 pieces, each
# of length 1 resulting in a revenue of 6 * 6 = 36.
_A : Any = 36
_A : List[Any] = top_down_cut_rod(snake_case_,snake_case_ )
_A : List[Any] = bottom_up_cut_rod(snake_case_,snake_case_ )
_A : Dict = naive_cut_rod_recursive(snake_case_,snake_case_ )
assert expected_max_revenue == max_rev_top_down
assert max_rev_top_down == max_rev_bottom_up
assert max_rev_bottom_up == max_rev_naive
if __name__ == "__main__":
main()
| 26 | 1 |
import collections
import os
import re
from pathlib import Path
_snake_case = "src/transformers"
# Matches is_xxx_available()
_snake_case = re.compile(r"is\_([a-z_]*)_available()")
# Catches a one-line _import_struct = {xxx}
_snake_case = re.compile(r"^_import_structure\s+=\s+\{([^\}]+)\}")
# Catches a line with a key-values pattern: "bla": ["foo", "bar"]
_snake_case = re.compile(r"\s+\"\S*\":\s+\[([^\]]*)\]")
# Catches a line if not is_foo_available
_snake_case = re.compile(r"^\s*if\s+not\s+is\_[a-z_]*\_available\(\)")
# Catches a line _import_struct["bla"].append("foo")
_snake_case = re.compile(r"^\s*_import_structure\[\"\S*\"\]\.append\(\"(\S*)\"\)")
# Catches a line _import_struct["bla"].extend(["foo", "bar"]) or _import_struct["bla"] = ["foo", "bar"]
_snake_case = re.compile(r"^\s*_import_structure\[\S*\](?:\.extend\(|\s*=\s+)\[([^\]]*)\]")
# Catches a line with an object between quotes and a comma: "MyModel",
_snake_case = re.compile(r"^\s+\"([^\"]+)\",")
# Catches a line with objects between brackets only: ["foo", "bar"],
_snake_case = re.compile(r"^\s+\[([^\]]+)\]")
# Catches a line with from foo import bar, bla, boo
_snake_case = re.compile(r"\s+from\s+\S*\s+import\s+([^\(\s].*)\n")
# Catches a line with try:
_snake_case = re.compile(r"^\s*try:")
# Catches a line with else:
_snake_case = re.compile(r"^\s*else:")
def lowerCAmelCase_ ( snake_case_ ):
if _re_test_backend.search(snake_case_ ) is None:
return None
_A : Any = [b[0] for b in _re_backend.findall(snake_case_ )]
backends.sort()
return "_and_".join(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
with open(snake_case_,"""r""",encoding="""utf-8""",newline="""\n""" ) as f:
_A : Dict = f.readlines()
_A : Union[str, Any] = 0
while line_index < len(snake_case_ ) and not lines[line_index].startswith("""_import_structure = {""" ):
line_index += 1
# If this is a traditional init, just return.
if line_index >= len(snake_case_ ):
return None
# First grab the objects without a specific backend in _import_structure
_A : List[Any] = []
while not lines[line_index].startswith("""if TYPE_CHECKING""" ) and find_backend(lines[line_index] ) is None:
_A : Dict = lines[line_index]
# If we have everything on a single line, let's deal with it.
if _re_one_line_import_struct.search(snake_case_ ):
_A : Dict = _re_one_line_import_struct.search(snake_case_ ).groups()[0]
_A : Any = re.findall(r"""\[([^\]]+)\]""",snake_case_ )
for imp in imports:
objects.extend([obj[1:-1] for obj in imp.split(""", """ )] )
line_index += 1
continue
_A : Optional[int] = _re_import_struct_key_value.search(snake_case_ )
if single_line_import_search is not None:
_A : Union[str, Any] = [obj[1:-1] for obj in single_line_import_search.groups()[0].split(""", """ ) if len(snake_case_ ) > 0]
objects.extend(snake_case_ )
elif line.startswith(""" """ * 8 + """\"""" ):
objects.append(line[9:-3] )
line_index += 1
_A : Any = {"""none""": objects}
# Let's continue with backend-specific objects in _import_structure
while not lines[line_index].startswith("""if TYPE_CHECKING""" ):
# If the line is an if not is_backend_available, we grab all objects associated.
_A : str = find_backend(lines[line_index] )
# Check if the backend declaration is inside a try block:
if _re_try.search(lines[line_index - 1] ) is None:
_A : Union[str, Any] = None
if backend is not None:
line_index += 1
# Scroll until we hit the else block of try-except-else
while _re_else.search(lines[line_index] ) is None:
line_index += 1
line_index += 1
_A : Tuple = []
# Until we unindent, add backend objects to the list
while len(lines[line_index] ) <= 1 or lines[line_index].startswith(""" """ * 4 ):
_A : List[str] = lines[line_index]
if _re_import_struct_add_one.search(snake_case_ ) is not None:
objects.append(_re_import_struct_add_one.search(snake_case_ ).groups()[0] )
elif _re_import_struct_add_many.search(snake_case_ ) is not None:
_A : str = _re_import_struct_add_many.search(snake_case_ ).groups()[0].split(""", """ )
_A : str = [obj[1:-1] for obj in imports if len(snake_case_ ) > 0]
objects.extend(snake_case_ )
elif _re_between_brackets.search(snake_case_ ) is not None:
_A : Dict = _re_between_brackets.search(snake_case_ ).groups()[0].split(""", """ )
_A : Dict = [obj[1:-1] for obj in imports if len(snake_case_ ) > 0]
objects.extend(snake_case_ )
elif _re_quote_object.search(snake_case_ ) is not None:
objects.append(_re_quote_object.search(snake_case_ ).groups()[0] )
elif line.startswith(""" """ * 8 + """\"""" ):
objects.append(line[9:-3] )
elif line.startswith(""" """ * 12 + """\"""" ):
objects.append(line[13:-3] )
line_index += 1
_A : Optional[Any] = objects
else:
line_index += 1
# At this stage we are in the TYPE_CHECKING part, first grab the objects without a specific backend
_A : str = []
while (
line_index < len(snake_case_ )
and find_backend(lines[line_index] ) is None
and not lines[line_index].startswith("""else""" )
):
_A : Dict = lines[line_index]
_A : List[Any] = _re_import.search(snake_case_ )
if single_line_import_search is not None:
objects.extend(single_line_import_search.groups()[0].split(""", """ ) )
elif line.startswith(""" """ * 8 ):
objects.append(line[8:-2] )
line_index += 1
_A : Optional[int] = {"""none""": objects}
# Let's continue with backend-specific objects
while line_index < len(snake_case_ ):
# If the line is an if is_backend_available, we grab all objects associated.
_A : Union[str, Any] = find_backend(lines[line_index] )
# Check if the backend declaration is inside a try block:
if _re_try.search(lines[line_index - 1] ) is None:
_A : List[Any] = None
if backend is not None:
line_index += 1
# Scroll until we hit the else block of try-except-else
while _re_else.search(lines[line_index] ) is None:
line_index += 1
line_index += 1
_A : Dict = []
# Until we unindent, add backend objects to the list
while len(lines[line_index] ) <= 1 or lines[line_index].startswith(""" """ * 8 ):
_A : Optional[int] = lines[line_index]
_A : str = _re_import.search(snake_case_ )
if single_line_import_search is not None:
objects.extend(single_line_import_search.groups()[0].split(""", """ ) )
elif line.startswith(""" """ * 12 ):
objects.append(line[12:-2] )
line_index += 1
_A : int = objects
else:
line_index += 1
return import_dict_objects, type_hint_objects
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
def find_duplicates(snake_case_ ):
return [k for k, v in collections.Counter(snake_case_ ).items() if v > 1]
if list(import_dict_objects.keys() ) != list(type_hint_objects.keys() ):
return ["Both sides of the init do not have the same backends!"]
_A : Tuple = []
for key in import_dict_objects.keys():
_A : List[str] = find_duplicates(import_dict_objects[key] )
if duplicate_imports:
errors.append(f'''Duplicate _import_structure definitions for: {duplicate_imports}''' )
_A : int = find_duplicates(type_hint_objects[key] )
if duplicate_type_hints:
errors.append(f'''Duplicate TYPE_CHECKING objects for: {duplicate_type_hints}''' )
if sorted(set(import_dict_objects[key] ) ) != sorted(set(type_hint_objects[key] ) ):
_A : int = """base imports""" if key == """none""" else f'''{key} backend'''
errors.append(f'''Differences for {name}:''' )
for a in type_hint_objects[key]:
if a not in import_dict_objects[key]:
errors.append(f''' {a} in TYPE_HINT but not in _import_structure.''' )
for a in import_dict_objects[key]:
if a not in type_hint_objects[key]:
errors.append(f''' {a} in _import_structure but not in TYPE_HINT.''' )
return errors
def lowerCAmelCase_ ( ):
_A : str = []
for root, _, files in os.walk(snake_case_ ):
if "__init__.py" in files:
_A : str = os.path.join(snake_case_,"""__init__.py""" )
_A : Any = parse_init(snake_case_ )
if objects is not None:
_A : Optional[int] = analyze_results(*snake_case_ )
if len(snake_case_ ) > 0:
_A : Optional[Any] = f'''Problem in {fname}, both halves do not define the same objects.\n{errors[0]}'''
failures.append("""\n""".join(snake_case_ ) )
if len(snake_case_ ) > 0:
raise ValueError("""\n\n""".join(snake_case_ ) )
def lowerCAmelCase_ ( ):
_A : Optional[Any] = []
for path, directories, files in os.walk(snake_case_ ):
for folder in directories:
# Ignore private modules
if folder.startswith("""_""" ):
directories.remove(snake_case_ )
continue
# Ignore leftovers from branches (empty folders apart from pycache)
if len(list((Path(snake_case_ ) / folder).glob("""*.py""" ) ) ) == 0:
continue
_A : Optional[int] = str((Path(snake_case_ ) / folder).relative_to(snake_case_ ) )
_A : Optional[int] = short_path.replace(os.path.sep,""".""" )
submodules.append(snake_case_ )
for fname in files:
if fname == "__init__.py":
continue
_A : Optional[Any] = str((Path(snake_case_ ) / fname).relative_to(snake_case_ ) )
_A : Dict = short_path.replace(""".py""","""""" ).replace(os.path.sep,""".""" )
if len(submodule.split(""".""" ) ) == 1:
submodules.append(snake_case_ )
return submodules
_snake_case = [
"convert_pytorch_checkpoint_to_tf2",
"modeling_flax_pytorch_utils",
"models.esm.openfold_utils",
]
def lowerCAmelCase_ ( ):
# This is to make sure the transformers module imported is the one in the repo.
from transformers.utils import direct_transformers_import
_A : List[Any] = direct_transformers_import(snake_case_ )
_A : List[str] = set(transformers._import_structure.keys() )
# This contains all the base keys of the _import_structure object defined in the init, but if the user is missing
# some optional dependencies, they may not have all of them. Thus we read the init to read all additions and
# (potentiall re-) add them.
with open(os.path.join(snake_case_,"""__init__.py""" ),"""r""" ) as f:
_A : str = f.read()
import_structure_keys.update(set(re.findall(r"""import_structure\[\"([^\"]*)\"\]""",snake_case_ ) ) )
_A : str = [
module
for module in get_transformers_submodules()
if module not in IGNORE_SUBMODULES and module not in import_structure_keys
]
if len(snake_case_ ) > 0:
_A : Union[str, Any] = """\n""".join(f'''- {module}''' for module in module_not_registered )
raise ValueError(
"""The following submodules are not properly registed in the main init of Transformers:\n"""
f'''{list_of_modules}\n'''
"""Make sure they appear somewhere in the keys of `_import_structure` with an empty list as value.""" )
if __name__ == "__main__":
check_all_inits()
check_submodules()
| 26 |
import requests
from bsa import BeautifulSoup
def lowerCAmelCase_ ( snake_case_ = "AAPL" ):
_A : str = f'''https://in.finance.yahoo.com/quote/{symbol}?s={symbol}'''
_A : List[Any] = BeautifulSoup(requests.get(snake_case_ ).text,"""html.parser""" )
_A : Union[str, Any] = """My(6px) Pos(r) smartphone_Mt(6px)"""
return soup.find("""div""",class_=class_ ).find("""span""" ).text
if __name__ == "__main__":
for symbol in "AAPL AMZN IBM GOOG MSFT ORCL".split():
print(f"""Current {symbol:<4} stock price is {stock_price(symbol):>8}""")
| 26 | 1 |
def lowerCAmelCase_ ( snake_case_ ):
_A : Optional[Any] = 0
for ch in input_str:
_A : List[str] = ord(snake_case_ )
_A : List[str] = pow(2,snake_case_ )
# If we already turned on bit for current character's unicode
if bitmap >> ch_unicode & 1 == 1:
return False
bitmap |= ch_bit_index_on
return True
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 |
import unittest
from transformers import (
MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
TextaTextGenerationPipeline,
pipeline,
)
from transformers.testing_utils import is_pipeline_test, require_tf, require_torch
from transformers.utils import is_torch_available
from .test_pipelines_common import ANY
if is_torch_available():
import torch
@is_pipeline_test
class lowercase ( unittest.TestCase ):
_a = MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING
_a = TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING
def a__ ( self , _a , _a , _a ) -> int:
_A : str = TextaTextGenerationPipeline(model=_a , tokenizer=_a )
return generator, ["Something to write", "Something else"]
def a__ ( self , _a , _a ) -> Dict:
_A : Any = generator("""Something there""" )
self.assertEqual(_a , [{"""generated_text""": ANY(_a )}] )
# These are encoder decoder, they don't just append to incoming string
self.assertFalse(outputs[0]["""generated_text"""].startswith("""Something there""" ) )
_A : List[Any] = generator(["""This is great !""", """Something else"""] , num_return_sequences=2 , do_sample=_a )
self.assertEqual(
_a , [
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
] , )
_A : Optional[int] = generator(
["""This is great !""", """Something else"""] , num_return_sequences=2 , batch_size=2 , do_sample=_a )
self.assertEqual(
_a , [
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
] , )
with self.assertRaises(_a ):
generator(4 )
@require_torch
def a__ ( self ) -> List[str]:
_A : Any = pipeline("""text2text-generation""" , model="""patrickvonplaten/t5-tiny-random""" , framework="""pt""" )
# do_sample=False necessary for reproducibility
_A : Dict = generator("""Something there""" , do_sample=_a )
self.assertEqual(_a , [{"""generated_text""": """"""}] )
_A : Any = 3
_A : Any = generator(
"""Something there""" , num_return_sequences=_a , num_beams=_a , )
_A : Optional[int] = [
{"""generated_text""": """Beide Beide Beide Beide Beide Beide Beide Beide Beide"""},
{"""generated_text""": """Beide Beide Beide Beide Beide Beide Beide Beide"""},
{"""generated_text""": """"""},
]
self.assertEqual(_a , _a )
_A : Dict = generator("""This is a test""" , do_sample=_a , num_return_sequences=2 , return_tensors=_a )
self.assertEqual(
_a , [
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
] , )
_A : Dict = generator.model.config.eos_token_id
_A : List[str] = """<pad>"""
_A : Dict = generator(
["""This is a test""", """This is a second test"""] , do_sample=_a , num_return_sequences=2 , batch_size=2 , return_tensors=_a , )
self.assertEqual(
_a , [
[
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
],
[
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
],
] , )
@require_tf
def a__ ( self ) -> int:
_A : Optional[Any] = pipeline("""text2text-generation""" , model="""patrickvonplaten/t5-tiny-random""" , framework="""tf""" )
# do_sample=False necessary for reproducibility
_A : str = generator("""Something there""" , do_sample=_a )
self.assertEqual(_a , [{"""generated_text""": """"""}] )
| 26 | 1 |
import unittest
from transformers import is_torch_available
from transformers.testing_utils import require_sentencepiece, require_tokenizers, require_torch, slow, torch_device
if is_torch_available():
from transformers import AutoModelForSeqaSeqLM, AutoTokenizer
@require_torch
@require_sentencepiece
@require_tokenizers
class lowercase ( unittest.TestCase ):
@slow
def a__ ( self ) -> Any:
_A : Dict = AutoModelForSeqaSeqLM.from_pretrained("""google/mt5-small""" , return_dict=_a ).to(_a )
_A : Tuple = AutoTokenizer.from_pretrained("""google/mt5-small""" )
_A : List[Any] = tokenizer("""Hello there""" , return_tensors="""pt""" ).input_ids
_A : Optional[int] = tokenizer("""Hi I am""" , return_tensors="""pt""" ).input_ids
_A : Dict = model(input_ids.to(_a ) , labels=labels.to(_a ) ).loss
_A : Tuple = -(labels.shape[-1] * loss.item())
_A : Union[str, Any] = -84.9127
self.assertTrue(abs(mtf_score - EXPECTED_SCORE ) < 1e-4 )
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
while b:
_A , _A : List[str] = b, a % b
return a
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
return a if b == 0 else euclidean_gcd_recursive(snake_case_,a % b )
def lowerCAmelCase_ ( ):
print(f'''euclidean_gcd(3, 5) = {euclidean_gcd(3,5 )}''' )
print(f'''euclidean_gcd(5, 3) = {euclidean_gcd(5,3 )}''' )
print(f'''euclidean_gcd(1, 3) = {euclidean_gcd(1,3 )}''' )
print(f'''euclidean_gcd(3, 6) = {euclidean_gcd(3,6 )}''' )
print(f'''euclidean_gcd(6, 3) = {euclidean_gcd(6,3 )}''' )
print(f'''euclidean_gcd_recursive(3, 5) = {euclidean_gcd_recursive(3,5 )}''' )
print(f'''euclidean_gcd_recursive(5, 3) = {euclidean_gcd_recursive(5,3 )}''' )
print(f'''euclidean_gcd_recursive(1, 3) = {euclidean_gcd_recursive(1,3 )}''' )
print(f'''euclidean_gcd_recursive(3, 6) = {euclidean_gcd_recursive(3,6 )}''' )
print(f'''euclidean_gcd_recursive(6, 3) = {euclidean_gcd_recursive(6,3 )}''' )
if __name__ == "__main__":
main()
| 26 | 1 |
import inspect
import tempfile
from collections import OrderedDict, UserDict
from collections.abc import MutableMapping
from contextlib import ExitStack, contextmanager
from dataclasses import fields
from enum import Enum
from typing import Any, ContextManager, List, Tuple
import numpy as np
from .import_utils import is_flax_available, is_tf_available, is_torch_available, is_torch_fx_proxy
if is_flax_available():
import jax.numpy as jnp
class lowercase ( UpperCamelCase__ ):
def __get__( self , _a , _a=None ) -> Optional[int]:
# See docs.python.org/3/howto/descriptor.html#properties
if obj is None:
return self
if self.fget is None:
raise AttributeError("""unreadable attribute""" )
_A : List[str] = """__cached_""" + self.fget.__name__
_A : Any = getattr(_a , _a , _a )
if cached is None:
_A : Dict = self.fget(_a )
setattr(_a , _a , _a )
return cached
def lowerCAmelCase_ ( snake_case_ ):
_A : Union[str, Any] = val.lower()
if val in {"y", "yes", "t", "true", "on", "1"}:
return 1
if val in {"n", "no", "f", "false", "off", "0"}:
return 0
raise ValueError(f'''invalid truth value {val!r}''' )
def lowerCAmelCase_ ( snake_case_ ):
if is_torch_fx_proxy(snake_case_ ):
return True
if is_torch_available():
import torch
if isinstance(snake_case_,torch.Tensor ):
return True
if is_tf_available():
import tensorflow as tf
if isinstance(snake_case_,tf.Tensor ):
return True
if is_flax_available():
import jax.numpy as jnp
from jax.core import Tracer
if isinstance(snake_case_,(jnp.ndarray, Tracer) ):
return True
return isinstance(snake_case_,np.ndarray )
def lowerCAmelCase_ ( snake_case_ ):
return isinstance(snake_case_,np.ndarray )
def lowerCAmelCase_ ( snake_case_ ):
return _is_numpy(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
import torch
return isinstance(snake_case_,torch.Tensor )
def lowerCAmelCase_ ( snake_case_ ):
return False if not is_torch_available() else _is_torch(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
import torch
return isinstance(snake_case_,torch.device )
def lowerCAmelCase_ ( snake_case_ ):
return False if not is_torch_available() else _is_torch_device(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
import torch
if isinstance(snake_case_,snake_case_ ):
if hasattr(snake_case_,snake_case_ ):
_A : Optional[int] = getattr(snake_case_,snake_case_ )
else:
return False
return isinstance(snake_case_,torch.dtype )
def lowerCAmelCase_ ( snake_case_ ):
return False if not is_torch_available() else _is_torch_dtype(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
import tensorflow as tf
return isinstance(snake_case_,tf.Tensor )
def lowerCAmelCase_ ( snake_case_ ):
return False if not is_tf_available() else _is_tensorflow(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
import tensorflow as tf
# the `is_symbolic_tensor` predicate is only available starting with TF 2.14
if hasattr(snake_case_,"""is_symbolic_tensor""" ):
return tf.is_symbolic_tensor(snake_case_ )
return type(snake_case_ ) == tf.Tensor
def lowerCAmelCase_ ( snake_case_ ):
return False if not is_tf_available() else _is_tf_symbolic_tensor(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
import jax.numpy as jnp # noqa: F811
return isinstance(snake_case_,jnp.ndarray )
def lowerCAmelCase_ ( snake_case_ ):
return False if not is_flax_available() else _is_jax(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
if isinstance(snake_case_,(dict, UserDict) ):
return {k: to_py_obj(snake_case_ ) for k, v in obj.items()}
elif isinstance(snake_case_,(list, tuple) ):
return [to_py_obj(snake_case_ ) for o in obj]
elif is_tf_tensor(snake_case_ ):
return obj.numpy().tolist()
elif is_torch_tensor(snake_case_ ):
return obj.detach().cpu().tolist()
elif is_jax_tensor(snake_case_ ):
return np.asarray(snake_case_ ).tolist()
elif isinstance(snake_case_,(np.ndarray, np.number) ): # tolist also works on 0d np arrays
return obj.tolist()
else:
return obj
def lowerCAmelCase_ ( snake_case_ ):
if isinstance(snake_case_,(dict, UserDict) ):
return {k: to_numpy(snake_case_ ) for k, v in obj.items()}
elif isinstance(snake_case_,(list, tuple) ):
return np.array(snake_case_ )
elif is_tf_tensor(snake_case_ ):
return obj.numpy()
elif is_torch_tensor(snake_case_ ):
return obj.detach().cpu().numpy()
elif is_jax_tensor(snake_case_ ):
return np.asarray(snake_case_ )
else:
return obj
class lowercase ( UpperCamelCase__ ):
def a__ ( self ) -> str:
_A : int = fields(self )
# Safety and consistency checks
if not len(_a ):
raise ValueError(F'''{self.__class__.__name__} has no fields.''' )
if not all(field.default is None for field in class_fields[1:] ):
raise ValueError(F'''{self.__class__.__name__} should not have more than one required field.''' )
_A : str = getattr(self , class_fields[0].name )
_A : str = all(getattr(self , field.name ) is None for field in class_fields[1:] )
if other_fields_are_none and not is_tensor(_a ):
if isinstance(_a , _a ):
_A : int = first_field.items()
_A : Dict = True
else:
try:
_A : List[Any] = iter(_a )
_A : List[Any] = True
except TypeError:
_A : Dict = False
# if we provided an iterator as first field and the iterator is a (key, value) iterator
# set the associated fields
if first_field_iterator:
for idx, element in enumerate(_a ):
if (
not isinstance(_a , (list, tuple) )
or not len(_a ) == 2
or not isinstance(element[0] , _a )
):
if idx == 0:
# If we do not have an iterator of key/values, set it as attribute
_A : Dict = first_field
else:
# If we have a mixed iterator, raise an error
raise ValueError(
F'''Cannot set key/value for {element}. It needs to be a tuple (key, value).''' )
break
setattr(self , element[0] , element[1] )
if element[1] is not None:
_A : Any = element[1]
elif first_field is not None:
_A : List[str] = first_field
else:
for field in class_fields:
_A : Any = getattr(self , field.name )
if v is not None:
_A : int = v
def __delitem__( self , *_a , **_a ) -> int:
raise Exception(F'''You cannot use ``__delitem__`` on a {self.__class__.__name__} instance.''' )
def a__ ( self , *_a , **_a ) -> Dict:
raise Exception(F'''You cannot use ``setdefault`` on a {self.__class__.__name__} instance.''' )
def a__ ( self , *_a , **_a ) -> Optional[Any]:
raise Exception(F'''You cannot use ``pop`` on a {self.__class__.__name__} instance.''' )
def a__ ( self , *_a , **_a ) -> List[Any]:
raise Exception(F'''You cannot use ``update`` on a {self.__class__.__name__} instance.''' )
def __getitem__( self , _a ) -> Optional[int]:
if isinstance(_a , _a ):
_A : str = dict(self.items() )
return inner_dict[k]
else:
return self.to_tuple()[k]
def __setattr__( self , _a , _a ) -> List[Any]:
if name in self.keys() and value is not None:
# Don't call self.__setitem__ to avoid recursion errors
super().__setitem__(_a , _a )
super().__setattr__(_a , _a )
def __setitem__( self , _a , _a ) -> int:
# Will raise a KeyException if needed
super().__setitem__(_a , _a )
# Don't call self.__setattr__ to avoid recursion errors
super().__setattr__(_a , _a )
def a__ ( self ) -> Tuple[Any]:
return tuple(self[k] for k in self.keys() )
class lowercase ( UpperCamelCase__,UpperCamelCase__ ):
@classmethod
def a__ ( cls , _a ) -> Any:
raise ValueError(
F'''{value} is not a valid {cls.__name__}, please select one of {list(cls._valueamember_map_.keys() )}''' )
class lowercase ( UpperCamelCase__ ):
_a = "longest"
_a = "max_length"
_a = "do_not_pad"
class lowercase ( UpperCamelCase__ ):
_a = "pt"
_a = "tf"
_a = "np"
_a = "jax"
class lowercase :
def __init__( self , _a ) -> Tuple:
_A : List[Any] = context_managers
_A : List[Any] = ExitStack()
def __enter__( self ) -> List[str]:
for context_manager in self.context_managers:
self.stack.enter_context(_a )
def __exit__( self , *_a , **_a ) -> Union[str, Any]:
self.stack.__exit__(*_a , **_a )
def lowerCAmelCase_ ( snake_case_ ):
_A : List[Any] = infer_framework(snake_case_ )
if framework == "tf":
_A : str = inspect.signature(model_class.call ) # TensorFlow models
elif framework == "pt":
_A : Union[str, Any] = inspect.signature(model_class.forward ) # PyTorch models
else:
_A : int = inspect.signature(model_class.__call__ ) # Flax models
for p in signature.parameters:
if p == "return_loss" and signature.parameters[p].default is True:
return True
return False
def lowerCAmelCase_ ( snake_case_ ):
_A : Dict = model_class.__name__
_A : Optional[Any] = infer_framework(snake_case_ )
if framework == "tf":
_A : Any = inspect.signature(model_class.call ) # TensorFlow models
elif framework == "pt":
_A : Optional[Any] = inspect.signature(model_class.forward ) # PyTorch models
else:
_A : Optional[Any] = inspect.signature(model_class.__call__ ) # Flax models
if "QuestionAnswering" in model_name:
return [p for p in signature.parameters if "label" in p or p in ("start_positions", "end_positions")]
else:
return [p for p in signature.parameters if "label" in p]
def lowerCAmelCase_ ( snake_case_,snake_case_ = "",snake_case_ = "." ):
def _flatten_dict(snake_case_,snake_case_="",snake_case_="." ):
for k, v in d.items():
_A : Dict = str(snake_case_ ) + delimiter + str(snake_case_ ) if parent_key else k
if v and isinstance(snake_case_,snake_case_ ):
yield from flatten_dict(snake_case_,snake_case_,delimiter=snake_case_ ).items()
else:
yield key, v
return dict(_flatten_dict(snake_case_,snake_case_,snake_case_ ) )
@contextmanager
def lowerCAmelCase_ ( snake_case_,snake_case_ = False ):
if use_temp_dir:
with tempfile.TemporaryDirectory() as tmp_dir:
yield tmp_dir
else:
yield working_dir
def lowerCAmelCase_ ( snake_case_,snake_case_=None ):
if is_numpy_array(snake_case_ ):
return np.transpose(snake_case_,axes=snake_case_ )
elif is_torch_tensor(snake_case_ ):
return array.T if axes is None else array.permute(*snake_case_ )
elif is_tf_tensor(snake_case_ ):
import tensorflow as tf
return tf.transpose(snake_case_,perm=snake_case_ )
elif is_jax_tensor(snake_case_ ):
return jnp.transpose(snake_case_,axes=snake_case_ )
else:
raise ValueError(f'''Type not supported for transpose: {type(snake_case_ )}.''' )
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if is_numpy_array(snake_case_ ):
return np.reshape(snake_case_,snake_case_ )
elif is_torch_tensor(snake_case_ ):
return array.reshape(*snake_case_ )
elif is_tf_tensor(snake_case_ ):
import tensorflow as tf
return tf.reshape(snake_case_,snake_case_ )
elif is_jax_tensor(snake_case_ ):
return jnp.reshape(snake_case_,snake_case_ )
else:
raise ValueError(f'''Type not supported for reshape: {type(snake_case_ )}.''' )
def lowerCAmelCase_ ( snake_case_,snake_case_=None ):
if is_numpy_array(snake_case_ ):
return np.squeeze(snake_case_,axis=snake_case_ )
elif is_torch_tensor(snake_case_ ):
return array.squeeze() if axis is None else array.squeeze(dim=snake_case_ )
elif is_tf_tensor(snake_case_ ):
import tensorflow as tf
return tf.squeeze(snake_case_,axis=snake_case_ )
elif is_jax_tensor(snake_case_ ):
return jnp.squeeze(snake_case_,axis=snake_case_ )
else:
raise ValueError(f'''Type not supported for squeeze: {type(snake_case_ )}.''' )
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if is_numpy_array(snake_case_ ):
return np.expand_dims(snake_case_,snake_case_ )
elif is_torch_tensor(snake_case_ ):
return array.unsqueeze(dim=snake_case_ )
elif is_tf_tensor(snake_case_ ):
import tensorflow as tf
return tf.expand_dims(snake_case_,axis=snake_case_ )
elif is_jax_tensor(snake_case_ ):
return jnp.expand_dims(snake_case_,axis=snake_case_ )
else:
raise ValueError(f'''Type not supported for expand_dims: {type(snake_case_ )}.''' )
def lowerCAmelCase_ ( snake_case_ ):
if is_numpy_array(snake_case_ ):
return np.size(snake_case_ )
elif is_torch_tensor(snake_case_ ):
return array.numel()
elif is_tf_tensor(snake_case_ ):
import tensorflow as tf
return tf.size(snake_case_ )
elif is_jax_tensor(snake_case_ ):
return array.size
else:
raise ValueError(f'''Type not supported for expand_dims: {type(snake_case_ )}.''' )
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
for key, value in auto_map.items():
if isinstance(snake_case_,(tuple, list) ):
_A : int = [f'''{repo_id}--{v}''' if (v is not None and """--""" not in v) else v for v in value]
elif value is not None and "--" not in value:
_A : List[str] = f'''{repo_id}--{value}'''
return auto_map
def lowerCAmelCase_ ( snake_case_ ):
for base_class in inspect.getmro(snake_case_ ):
_A : Optional[Any] = base_class.__module__
_A : Tuple = base_class.__name__
if module.startswith("""tensorflow""" ) or module.startswith("""keras""" ) or name == "TFPreTrainedModel":
return "tf"
elif module.startswith("""torch""" ) or name == "PreTrainedModel":
return "pt"
elif module.startswith("""flax""" ) or module.startswith("""jax""" ) or name == "FlaxPreTrainedModel":
return "flax"
else:
raise TypeError(f'''Could not infer framework from class {model_class}.''' )
| 26 |
def lowerCAmelCase_ ( snake_case_ ):
if number < 0:
raise ValueError("""number must not be negative""" )
return number & (number - 1) == 0
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
from dataclasses import dataclass
from typing import Optional
import numpy as np
import torch
import torch.nn as nn
from ..utils import BaseOutput, is_torch_version, randn_tensor
from .attention_processor import SpatialNorm
from .unet_ad_blocks import UNetMidBlockaD, get_down_block, get_up_block
@dataclass
class lowercase ( UpperCamelCase__ ):
_a = 42
class lowercase ( nn.Module ):
def __init__( self , _a=3 , _a=3 , _a=("DownEncoderBlock2D",) , _a=(64,) , _a=2 , _a=32 , _a="silu" , _a=True , ) -> Any:
super().__init__()
_A : List[str] = layers_per_block
_A : Optional[Any] = torch.nn.Convad(
_a , block_out_channels[0] , kernel_size=3 , stride=1 , padding=1 , )
_A : str = None
_A : int = nn.ModuleList([] )
# down
_A : Tuple = block_out_channels[0]
for i, down_block_type in enumerate(_a ):
_A : int = output_channel
_A : List[str] = block_out_channels[i]
_A : Optional[Any] = i == len(_a ) - 1
_A : Tuple = get_down_block(
_a , num_layers=self.layers_per_block , in_channels=_a , out_channels=_a , add_downsample=not is_final_block , resnet_eps=1e-6 , downsample_padding=0 , resnet_act_fn=_a , resnet_groups=_a , attention_head_dim=_a , temb_channels=_a , )
self.down_blocks.append(_a )
# mid
_A : List[Any] = UNetMidBlockaD(
in_channels=block_out_channels[-1] , resnet_eps=1e-6 , resnet_act_fn=_a , output_scale_factor=1 , resnet_time_scale_shift="""default""" , attention_head_dim=block_out_channels[-1] , resnet_groups=_a , temb_channels=_a , )
# out
_A : Any = nn.GroupNorm(num_channels=block_out_channels[-1] , num_groups=_a , eps=1e-6 )
_A : Optional[Any] = nn.SiLU()
_A : List[str] = 2 * out_channels if double_z else out_channels
_A : List[Any] = nn.Convad(block_out_channels[-1] , _a , 3 , padding=1 )
_A : Any = False
def a__ ( self , _a ) -> List[str]:
_A : Any = x
_A : Optional[Any] = self.conv_in(_a )
if self.training and self.gradient_checkpointing:
def create_custom_forward(_a ):
def custom_forward(*_a ):
return module(*_a )
return custom_forward
# down
if is_torch_version(""">=""" , """1.11.0""" ):
for down_block in self.down_blocks:
_A : int = torch.utils.checkpoint.checkpoint(
create_custom_forward(_a ) , _a , use_reentrant=_a )
# middle
_A : str = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.mid_block ) , _a , use_reentrant=_a )
else:
for down_block in self.down_blocks:
_A : str = torch.utils.checkpoint.checkpoint(create_custom_forward(_a ) , _a )
# middle
_A : Any = torch.utils.checkpoint.checkpoint(create_custom_forward(self.mid_block ) , _a )
else:
# down
for down_block in self.down_blocks:
_A : List[Any] = down_block(_a )
# middle
_A : Dict = self.mid_block(_a )
# post-process
_A : Union[str, Any] = self.conv_norm_out(_a )
_A : List[str] = self.conv_act(_a )
_A : str = self.conv_out(_a )
return sample
class lowercase ( nn.Module ):
def __init__( self , _a=3 , _a=3 , _a=("UpDecoderBlock2D",) , _a=(64,) , _a=2 , _a=32 , _a="silu" , _a="group" , ) -> Optional[int]:
super().__init__()
_A : Tuple = layers_per_block
_A : Any = nn.Convad(
_a , block_out_channels[-1] , kernel_size=3 , stride=1 , padding=1 , )
_A : Optional[Any] = None
_A : str = nn.ModuleList([] )
_A : Optional[int] = in_channels if norm_type == """spatial""" else None
# mid
_A : List[str] = UNetMidBlockaD(
in_channels=block_out_channels[-1] , resnet_eps=1e-6 , resnet_act_fn=_a , output_scale_factor=1 , resnet_time_scale_shift="""default""" if norm_type == """group""" else norm_type , attention_head_dim=block_out_channels[-1] , resnet_groups=_a , temb_channels=_a , )
# up
_A : List[str] = list(reversed(_a ) )
_A : List[Any] = reversed_block_out_channels[0]
for i, up_block_type in enumerate(_a ):
_A : Any = output_channel
_A : List[str] = reversed_block_out_channels[i]
_A : Optional[Any] = i == len(_a ) - 1
_A : List[str] = get_up_block(
_a , num_layers=self.layers_per_block + 1 , in_channels=_a , out_channels=_a , prev_output_channel=_a , add_upsample=not is_final_block , resnet_eps=1e-6 , resnet_act_fn=_a , resnet_groups=_a , attention_head_dim=_a , temb_channels=_a , resnet_time_scale_shift=_a , )
self.up_blocks.append(_a )
_A : Optional[int] = output_channel
# out
if norm_type == "spatial":
_A : Dict = SpatialNorm(block_out_channels[0] , _a )
else:
_A : List[str] = nn.GroupNorm(num_channels=block_out_channels[0] , num_groups=_a , eps=1e-6 )
_A : List[str] = nn.SiLU()
_A : str = nn.Convad(block_out_channels[0] , _a , 3 , padding=1 )
_A : int = False
def a__ ( self , _a , _a=None ) -> str:
_A : Union[str, Any] = z
_A : Dict = self.conv_in(_a )
_A : str = next(iter(self.up_blocks.parameters() ) ).dtype
if self.training and self.gradient_checkpointing:
def create_custom_forward(_a ):
def custom_forward(*_a ):
return module(*_a )
return custom_forward
if is_torch_version(""">=""" , """1.11.0""" ):
# middle
_A : str = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.mid_block ) , _a , _a , use_reentrant=_a )
_A : Dict = sample.to(_a )
# up
for up_block in self.up_blocks:
_A : Tuple = torch.utils.checkpoint.checkpoint(
create_custom_forward(_a ) , _a , _a , use_reentrant=_a )
else:
# middle
_A : Optional[int] = torch.utils.checkpoint.checkpoint(
create_custom_forward(self.mid_block ) , _a , _a )
_A : Optional[int] = sample.to(_a )
# up
for up_block in self.up_blocks:
_A : Tuple = torch.utils.checkpoint.checkpoint(create_custom_forward(_a ) , _a , _a )
else:
# middle
_A : Tuple = self.mid_block(_a , _a )
_A : Dict = sample.to(_a )
# up
for up_block in self.up_blocks:
_A : Dict = up_block(_a , _a )
# post-process
if latent_embeds is None:
_A : List[Any] = self.conv_norm_out(_a )
else:
_A : int = self.conv_norm_out(_a , _a )
_A : Tuple = self.conv_act(_a )
_A : Tuple = self.conv_out(_a )
return sample
class lowercase ( nn.Module ):
def __init__( self , _a , _a , _a , _a=None , _a="random" , _a=False , _a=True ) -> Union[str, Any]:
super().__init__()
_A : List[str] = n_e
_A : List[str] = vq_embed_dim
_A : int = beta
_A : Optional[Any] = legacy
_A : int = nn.Embedding(self.n_e , self.vq_embed_dim )
self.embedding.weight.data.uniform_(-1.0 / self.n_e , 1.0 / self.n_e )
_A : Tuple = remap
if self.remap is not None:
self.register_buffer("""used""" , torch.tensor(np.load(self.remap ) ) )
_A : Optional[Any] = self.used.shape[0]
_A : Dict = unknown_index # "random" or "extra" or integer
if self.unknown_index == "extra":
_A : List[str] = self.re_embed
_A : str = self.re_embed + 1
print(
F'''Remapping {self.n_e} indices to {self.re_embed} indices. '''
F'''Using {self.unknown_index} for unknown indices.''' )
else:
_A : Optional[Any] = n_e
_A : str = sane_index_shape
def a__ ( self , _a ) -> List[str]:
_A : Union[str, Any] = inds.shape
assert len(_a ) > 1
_A : Dict = inds.reshape(ishape[0] , -1 )
_A : Any = self.used.to(_a )
_A : Union[str, Any] = (inds[:, :, None] == used[None, None, ...]).long()
_A : Optional[int] = match.argmax(-1 )
_A : Dict = match.sum(2 ) < 1
if self.unknown_index == "random":
_A : Dict = torch.randint(0 , self.re_embed , size=new[unknown].shape ).to(device=new.device )
else:
_A : Tuple = self.unknown_index
return new.reshape(_a )
def a__ ( self , _a ) -> Optional[int]:
_A : Dict = inds.shape
assert len(_a ) > 1
_A : Union[str, Any] = inds.reshape(ishape[0] , -1 )
_A : List[Any] = self.used.to(_a )
if self.re_embed > self.used.shape[0]: # extra token
_A : Union[str, Any] = 0 # simply set to zero
_A : int = torch.gather(used[None, :][inds.shape[0] * [0], :] , 1 , _a )
return back.reshape(_a )
def a__ ( self , _a ) -> str:
# reshape z -> (batch, height, width, channel) and flatten
_A : Dict = z.permute(0 , 2 , 3 , 1 ).contiguous()
_A : List[str] = z.view(-1 , self.vq_embed_dim )
# distances from z to embeddings e_j (z - e)^2 = z^2 + e^2 - 2 e * z
_A : Tuple = torch.argmin(torch.cdist(_a , self.embedding.weight ) , dim=1 )
_A : List[Any] = self.embedding(_a ).view(z.shape )
_A : Optional[int] = None
_A : int = None
# compute loss for embedding
if not self.legacy:
_A : Any = self.beta * torch.mean((z_q.detach() - z) ** 2 ) + torch.mean((z_q - z.detach()) ** 2 )
else:
_A : Union[str, Any] = torch.mean((z_q.detach() - z) ** 2 ) + self.beta * torch.mean((z_q - z.detach()) ** 2 )
# preserve gradients
_A : int = z + (z_q - z).detach()
# reshape back to match original input shape
_A : Dict = z_q.permute(0 , 3 , 1 , 2 ).contiguous()
if self.remap is not None:
_A : Dict = min_encoding_indices.reshape(z.shape[0] , -1 ) # add batch axis
_A : List[Any] = self.remap_to_used(_a )
_A : str = min_encoding_indices.reshape(-1 , 1 ) # flatten
if self.sane_index_shape:
_A : List[Any] = min_encoding_indices.reshape(z_q.shape[0] , z_q.shape[2] , z_q.shape[3] )
return z_q, loss, (perplexity, min_encodings, min_encoding_indices)
def a__ ( self , _a , _a ) -> Dict:
# shape specifying (batch, height, width, channel)
if self.remap is not None:
_A : Optional[int] = indices.reshape(shape[0] , -1 ) # add batch axis
_A : str = self.unmap_to_all(_a )
_A : Optional[Any] = indices.reshape(-1 ) # flatten again
# get quantized latent vectors
_A : Optional[int] = self.embedding(_a )
if shape is not None:
_A : Tuple = z_q.view(_a )
# reshape back to match original input shape
_A : str = z_q.permute(0 , 3 , 1 , 2 ).contiguous()
return z_q
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a=False ) -> List[str]:
_A : str = parameters
_A , _A : Dict = torch.chunk(_a , 2 , dim=1 )
_A : List[Any] = torch.clamp(self.logvar , -30.0 , 20.0 )
_A : Optional[int] = deterministic
_A : List[str] = torch.exp(0.5 * self.logvar )
_A : Dict = torch.exp(self.logvar )
if self.deterministic:
_A : str = torch.zeros_like(
self.mean , device=self.parameters.device , dtype=self.parameters.dtype )
def a__ ( self , _a = None ) -> torch.FloatTensor:
# make sure sample is on the same device as the parameters and has same dtype
_A : Optional[int] = randn_tensor(
self.mean.shape , generator=_a , device=self.parameters.device , dtype=self.parameters.dtype )
_A : Optional[Any] = self.mean + self.std * sample
return x
def a__ ( self , _a=None ) -> int:
if self.deterministic:
return torch.Tensor([0.0] )
else:
if other is None:
return 0.5 * torch.sum(torch.pow(self.mean , 2 ) + self.var - 1.0 - self.logvar , dim=[1, 2, 3] )
else:
return 0.5 * torch.sum(
torch.pow(self.mean - other.mean , 2 ) / other.var
+ self.var / other.var
- 1.0
- self.logvar
+ other.logvar , dim=[1, 2, 3] , )
def a__ ( self , _a , _a=[1, 2, 3] ) -> Dict:
if self.deterministic:
return torch.Tensor([0.0] )
_A : Dict = np.log(2.0 * np.pi )
return 0.5 * torch.sum(logtwopi + self.logvar + torch.pow(sample - self.mean , 2 ) / self.var , dim=_a )
def a__ ( self ) -> int:
return self.mean
| 26 |
import argparse
import torch
from transformers import BlenderbotConfig, BlenderbotForConditionalGeneration
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
_snake_case = [
["attention", "attn"],
["encoder_attention", "encoder_attn"],
["q_lin", "q_proj"],
["k_lin", "k_proj"],
["v_lin", "v_proj"],
["out_lin", "out_proj"],
["norm_embeddings", "layernorm_embedding"],
["position_embeddings", "embed_positions"],
["embeddings", "embed_tokens"],
["ffn.lin", "fc"],
]
def lowerCAmelCase_ ( snake_case_ ):
if k == "embeddings.weight":
return "shared.weight"
for parlai_name, hf_name in PATTERNS:
_A : str = k.replace(snake_case_,snake_case_ )
if k.startswith("""encoder""" ):
_A : Optional[Any] = k.replace(""".attn""",""".self_attn""" )
_A : Dict = k.replace("""norm1""","""self_attn_layer_norm""" )
_A : Optional[Any] = k.replace("""norm2""","""final_layer_norm""" )
elif k.startswith("""decoder""" ):
_A : str = k.replace("""norm1""","""self_attn_layer_norm""" )
_A : Any = k.replace("""norm2""","""encoder_attn_layer_norm""" )
_A : Optional[int] = k.replace("""norm3""","""final_layer_norm""" )
return k
def lowerCAmelCase_ ( snake_case_ ):
_A : List[Any] = [
"""model.encoder.layernorm_embedding.weight""",
"""model.encoder.layernorm_embedding.bias""",
"""model.decoder.layernorm_embedding.weight""",
"""model.decoder.layernorm_embedding.bias""",
]
for k in keys:
_A : str = sd.pop(snake_case_ )
_A : Optional[int] = k.replace("""layernorm_embedding""","""layer_norm""" )
assert new_k not in sd
_A : Optional[int] = v
_snake_case = ["START"]
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
_A : List[Any] = model["""model"""]
_A : Optional[Any] = BlenderbotConfig.from_json_file(snake_case_ )
_A : List[str] = BlenderbotForConditionalGeneration(snake_case_ )
_A : Tuple = m.model.state_dict().keys()
_A : Any = []
_A : Dict = {}
for k, v in sd.items():
if k in IGNORE_KEYS:
continue
_A : Optional[int] = rename_state_dict_key(snake_case_ )
if new_k not in valid_keys:
failures.append([k, new_k] )
else:
_A : Dict = v
if cfg.normalize_before: # Blenderbot-3B checkpoints. Rename layernorm_embedding -> layer_norm
rename_layernorm_keys(snake_case_ )
m.model.load_state_dict(snake_case_,strict=snake_case_ )
m.half()
m.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument("--src_path", type=str, help="like blenderbot-model.bin")
parser.add_argument("--save_dir", default="hf_blenderbot", type=str, help="Where to save converted model.")
parser.add_argument(
"--hf_config_json", default="blenderbot-3b-config.json", type=str, help="Path to config to use"
)
_snake_case = parser.parse_args()
convert_parlai_checkpoint(args.src_path, args.save_dir, args.hf_config_json)
| 26 | 1 |
# Logistic Regression from scratch
# In[62]:
# In[63]:
# importing all the required libraries
import numpy as np
from matplotlib import pyplot as plt
from sklearn import datasets
def lowerCAmelCase_ ( snake_case_ ):
return 1 / (1 + np.exp(-z ))
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
return (-y * np.log(snake_case_ ) - (1 - y) * np.log(1 - h )).mean()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Optional[Any] = np.dot(snake_case_,snake_case_ )
return np.sum(y * scores - np.log(1 + np.exp(snake_case_ ) ) )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_=70000 ):
_A : Optional[Any] = np.zeros(x.shape[1] )
for iterations in range(snake_case_ ):
_A : Any = np.dot(snake_case_,snake_case_ )
_A : Optional[int] = sigmoid_function(snake_case_ )
_A : List[Any] = np.dot(x.T,h - y ) / y.size
_A : Any = theta - alpha * gradient # updating the weights
_A : List[Any] = np.dot(snake_case_,snake_case_ )
_A : List[Any] = sigmoid_function(snake_case_ )
_A : List[str] = cost_function(snake_case_,snake_case_ )
if iterations % 100 == 0:
print(f'''loss: {j} \t''' ) # printing the loss after every 100 iterations
return theta
# In[68]:
if __name__ == "__main__":
_snake_case = datasets.load_iris()
_snake_case = iris.data[:, :2]
_snake_case = (iris.target != 0) * 1
_snake_case = 0.1
_snake_case = logistic_reg(alpha, x, y, max_iterations=70000)
print("theta: ", theta) # printing the theta i.e our weights vector
def lowerCAmelCase_ ( snake_case_ ):
return sigmoid_function(
np.dot(snake_case_,snake_case_ ) ) # predicting the value of probability from the logistic regression algorithm
plt.figure(figsize=(10, 6))
plt.scatter(x[y == 0][:, 0], x[y == 0][:, 1], color="b", label="0")
plt.scatter(x[y == 1][:, 0], x[y == 1][:, 1], color="r", label="1")
((_snake_case) , (_snake_case)) = (x[:, 0].min(), x[:, 0].max())
((_snake_case) , (_snake_case)) = (x[:, 1].min(), x[:, 1].max())
((_snake_case) , (_snake_case)) = np.meshgrid(np.linspace(xa_min, xa_max), np.linspace(xa_min, xa_max))
_snake_case = np.c_[xxa.ravel(), xxa.ravel()]
_snake_case = predict_prob(grid).reshape(xxa.shape)
plt.contour(xxa, xxa, probs, [0.5], linewidths=1, colors="black")
plt.legend()
plt.show()
| 26 |
import multiprocessing
from typing import TYPE_CHECKING, Optional, Union
from .. import Dataset, Features, config
from ..formatting import query_table
from ..packaged_modules.sql.sql import Sql
from ..utils import logging
from .abc import AbstractDatasetInputStream
if TYPE_CHECKING:
import sqlitea
import sqlalchemy
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a = None , _a = None , _a = False , **_a , ) -> int:
super().__init__(features=_a , cache_dir=_a , keep_in_memory=_a , **_a )
_A : Optional[int] = Sql(
cache_dir=_a , features=_a , sql=_a , con=_a , **_a , )
def a__ ( self ) -> Optional[Any]:
_A : Tuple = None
_A : int = None
_A : Tuple = None
_A : Union[str, Any] = None
self.builder.download_and_prepare(
download_config=_a , download_mode=_a , verification_mode=_a , base_path=_a , )
# Build dataset for splits
_A : int = self.builder.as_dataset(
split="""train""" , verification_mode=_a , in_memory=self.keep_in_memory )
return dataset
class lowercase :
def __init__( self , _a , _a , _a , _a = None , _a = None , **_a , ) -> Union[str, Any]:
if num_proc is not None and num_proc <= 0:
raise ValueError(F'''num_proc {num_proc} must be an integer > 0.''' )
_A : Dict = dataset
_A : int = name
_A : Union[str, Any] = con
_A : str = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE
_A : str = num_proc
_A : Optional[Any] = to_sql_kwargs
def a__ ( self ) -> int:
_A : Any = self.to_sql_kwargs.pop("""sql""" , _a )
_A : List[str] = self.to_sql_kwargs.pop("""con""" , _a )
_A : int = self.to_sql_kwargs.pop("""index""" , _a )
_A : List[str] = self._write(index=_a , **self.to_sql_kwargs )
return written
def a__ ( self , _a ) -> Optional[int]:
_A , _A , _A : List[str] = args
_A : int = {**to_sql_kwargs, """if_exists""": """append"""} if offset > 0 else to_sql_kwargs
_A : str = query_table(
table=self.dataset.data , key=slice(_a , offset + self.batch_size ) , indices=self.dataset._indices , )
_A : Tuple = batch.to_pandas()
_A : Union[str, Any] = df.to_sql(self.name , self.con , index=_a , **_a )
return num_rows or len(_a )
def a__ ( self , _a , **_a ) -> int:
_A : Any = 0
if self.num_proc is None or self.num_proc == 1:
for offset in logging.tqdm(
range(0 , len(self.dataset ) , self.batch_size ) , unit="""ba""" , disable=not logging.is_progress_bar_enabled() , desc="""Creating SQL from Arrow format""" , ):
written += self._batch_sql((offset, index, to_sql_kwargs) )
else:
_A , _A : Tuple = len(self.dataset ), self.batch_size
with multiprocessing.Pool(self.num_proc ) as pool:
for num_rows in logging.tqdm(
pool.imap(
self._batch_sql , [(offset, index, to_sql_kwargs) for offset in range(0 , _a , _a )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit="""ba""" , disable=not logging.is_progress_bar_enabled() , desc="""Creating SQL from Arrow format""" , ):
written += num_rows
return written
| 26 | 1 |
def lowerCAmelCase_ ( snake_case_ ):
_A : str = [0 for i in range(len(snake_case_ ) )]
# initialize interval's left pointer and right pointer
_A , _A : Any = 0, 0
for i in range(1,len(snake_case_ ) ):
# case when current index is inside the interval
if i <= right_pointer:
_A : str = min(right_pointer - i + 1,z_result[i - left_pointer] )
_A : Optional[int] = min_edge
while go_next(snake_case_,snake_case_,snake_case_ ):
z_result[i] += 1
# if new index's result gives us more right interval,
# we've to update left_pointer and right_pointer
if i + z_result[i] - 1 > right_pointer:
_A , _A : Tuple = i, i + z_result[i] - 1
return z_result
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
return i + z_result[i] < len(snake_case_ ) and s[z_result[i]] == s[i + z_result[i]]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : List[Any] = 0
# concatenate 'pattern' and 'input_str' and call z_function
# with concatenated string
_A : Optional[Any] = z_function(pattern + input_str )
for val in z_result:
# if value is greater then length of the pattern string
# that means this index is starting position of substring
# which is equal to pattern string
if val >= len(snake_case_ ):
answer += 1
return answer
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 |
from ...configuration_utils import PretrainedConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/config.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/config.json"
# See all FNet models at https://huggingface.co/models?filter=fnet
}
class lowercase ( UpperCamelCase__ ):
_a = "fnet"
def __init__( self , _a=3_2000 , _a=768 , _a=12 , _a=3072 , _a="gelu_new" , _a=0.1 , _a=512 , _a=4 , _a=0.02 , _a=1e-12 , _a=False , _a=512 , _a=3 , _a=1 , _a=2 , **_a , ) -> int:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Any = vocab_size
_A : str = max_position_embeddings
_A : Optional[Any] = hidden_size
_A : List[str] = num_hidden_layers
_A : List[str] = intermediate_size
_A : List[Any] = hidden_act
_A : List[str] = hidden_dropout_prob
_A : List[str] = initializer_range
_A : List[Any] = type_vocab_size
_A : List[Any] = layer_norm_eps
_A : List[str] = use_tpu_fourier_optimizations
_A : str = tpu_short_seq_length
| 26 | 1 |
import collections
import json
import math
import os
import re
import time
from fnmatch import fnmatch
from typing import Dict
import requests
from slack_sdk import WebClient
_snake_case = WebClient(token=os.environ["CI_SLACK_BOT_TOKEN"])
def lowerCAmelCase_ ( snake_case_ ):
_A : Any = test_results.split(""" """ )
_A : int = 0
_A : Union[str, Any] = 0
# When the output is short enough, the output is surrounded by = signs: "== OUTPUT =="
# When it is too long, those signs are not present.
_A : Union[str, Any] = expressions[-2] if """=""" in expressions[-1] else expressions[-1]
for i, expression in enumerate(snake_case_ ):
if "failed" in expression:
failed += int(expressions[i - 1] )
if "passed" in expression:
success += int(expressions[i - 1] )
return failed, success, time_spent
def lowerCAmelCase_ ( snake_case_ ):
_A : Optional[Any] = {}
_A : Optional[int] = None
_A : Union[str, Any] = False
for line in failures_short_lines.split("""\n""" ):
if re.search(r"""_ \[doctest\]""",snake_case_ ):
_A : Any = True
_A : Tuple = line.split(""" """ )[2]
elif in_error and not line.split(""" """ )[0].isdigit():
_A : List[str] = line
_A : Optional[int] = False
return failures
class lowercase :
def __init__( self , _a , _a ) -> str:
_A : int = title
_A : List[str] = doc_test_results["""time_spent"""].split(""",""" )[0]
_A : Union[str, Any] = doc_test_results["""success"""]
_A : str = doc_test_results["""failures"""]
_A : Optional[Any] = self.n_success + self.n_failures
# Failures and success of the modeling tests
_A : Optional[int] = doc_test_results
@property
def a__ ( self ) -> str:
_A : str = [self._time_spent]
_A : Dict = 0
for time in time_spent:
_A : List[Any] = time.split(""":""" )
# Time can be formatted as xx:xx:xx, as .xx, or as x.xx if the time spent was less than a minute.
if len(_a ) == 1:
_A : Any = [0, 0, time_parts[0]]
_A , _A , _A : Tuple = int(time_parts[0] ), int(time_parts[1] ), float(time_parts[2] )
total_secs += hours * 3600 + minutes * 60 + seconds
_A , _A , _A : int = total_secs // 3600, (total_secs % 3600) // 60, total_secs % 60
return F'''{int(_a )}h{int(_a )}m{int(_a )}s'''
@property
def a__ ( self ) -> Dict:
return {"type": "header", "text": {"type": "plain_text", "text": self.title}}
@property
def a__ ( self ) -> Dict:
return {
"type": "section",
"text": {
"type": "plain_text",
"text": F'''🌞 There were no failures: all {self.n_tests} tests passed. The suite ran in {self.time}.''',
"emoji": True,
},
"accessory": {
"type": "button",
"text": {"type": "plain_text", "text": "Check Action results", "emoji": True},
"url": F'''https://github.com/huggingface/transformers/actions/runs/{os.environ["GITHUB_RUN_ID"]}''',
},
}
@property
def a__ ( self ) -> Dict:
return {
"type": "section",
"text": {
"type": "plain_text",
"text": (
F'''There were {self.n_failures} failures, out of {self.n_tests} tests.\nThe suite ran in'''
F''' {self.time}.'''
),
"emoji": True,
},
"accessory": {
"type": "button",
"text": {"type": "plain_text", "text": "Check Action results", "emoji": True},
"url": F'''https://github.com/huggingface/transformers/actions/runs/{os.environ["GITHUB_RUN_ID"]}''',
},
}
@property
def a__ ( self ) -> Dict:
_A : int = 40
_A : int = {k: v["""failed"""] for k, v in doc_test_results.items() if isinstance(_a , _a )}
_A : int = """"""
for category, failures in category_failures.items():
if len(_a ) == 0:
continue
if report != "":
report += "\n\n"
report += F'''*{category} failures*:'''.ljust(line_length // 2 ).rjust(line_length // 2 ) + "\n"
report += "`"
report += "`\n`".join(_a )
report += "`"
return {
"type": "section",
"text": {
"type": "mrkdwn",
"text": F'''The following examples had failures:\n\n\n{report}\n''',
},
}
@property
def a__ ( self ) -> str:
_A : List[Any] = [self.header]
if self.n_failures > 0:
blocks.append(self.failures )
if self.n_failures > 0:
blocks.extend([self.category_failures] )
if self.n_failures == 0:
blocks.append(self.no_failures )
return json.dumps(_a )
@staticmethod
def a__ ( ) -> int:
_A : List[str] = [
{
"""type""": """section""",
"""text""": {
"""type""": """plain_text""",
"""text""": """There was an issue running the tests.""",
},
"""accessory""": {
"""type""": """button""",
"""text""": {"""type""": """plain_text""", """text""": """Check Action results""", """emoji""": True},
"""url""": F'''https://github.com/huggingface/transformers/actions/runs/{os.environ["GITHUB_RUN_ID"]}''',
},
}
]
print("""Sending the following payload""" )
print(json.dumps({"""blocks""": json.loads(_a )} ) )
client.chat_postMessage(
channel=os.environ["""CI_SLACK_CHANNEL_ID_DAILY"""] , text="""There was an issue running the tests.""" , blocks=_a , )
def a__ ( self ) -> Optional[int]:
print("""Sending the following payload""" )
print(json.dumps({"""blocks""": json.loads(self.payload )} ) )
_A : List[Any] = F'''{self.n_failures} failures out of {self.n_tests} tests,''' if self.n_failures else """All tests passed."""
_A : Optional[int] = client.chat_postMessage(
channel=os.environ["""CI_SLACK_CHANNEL_ID_DAILY"""] , blocks=self.payload , text=_a , )
def a__ ( self , _a , _a , _a , _a ) -> Union[str, Any]:
_A : List[Any] = """"""
for key, value in failures.items():
_A : Dict = value[:200] + """ [Truncated]""" if len(_a ) > 250 else value
failures_text += F'''*{key}*\n_{value}_\n\n'''
_A : int = job_name
_A : int = {"""type""": """section""", """text""": {"""type""": """mrkdwn""", """text""": text}}
if job_link is not None:
_A : List[str] = {
"""type""": """button""",
"""text""": {"""type""": """plain_text""", """text""": """GitHub Action job""", """emoji""": True},
"""url""": job_link,
}
return [
{"type": "header", "text": {"type": "plain_text", "text": title.upper(), "emoji": True}},
content,
{"type": "section", "text": {"type": "mrkdwn", "text": failures_text}},
]
def a__ ( self ) -> Tuple:
if self.thread_ts is None:
raise ValueError("""Can only post reply if a post has been made.""" )
_A : List[str] = self.doc_test_results.pop("""job_link""" )
self.doc_test_results.pop("""failures""" )
self.doc_test_results.pop("""success""" )
self.doc_test_results.pop("""time_spent""" )
_A : Dict = sorted(self.doc_test_results.items() , key=lambda _a : t[0] )
for job, job_result in sorted_dict:
if len(job_result["""failures"""] ):
_A : Optional[int] = F'''*Num failures* :{len(job_result["failed"] )} \n'''
_A : Dict = job_result["""failures"""]
_A : Dict = self.get_reply_blocks(_a , _a , _a , text=_a )
print("""Sending the following reply""" )
print(json.dumps({"""blocks""": blocks} ) )
client.chat_postMessage(
channel=os.environ["""CI_SLACK_CHANNEL_ID_DAILY"""] , text=F'''Results for {job}''' , blocks=_a , thread_ts=self.thread_ts["""ts"""] , )
time.sleep(1 )
def lowerCAmelCase_ ( ):
_A : Dict = os.environ["""GITHUB_RUN_ID"""]
_A : Optional[Any] = f'''https://api.github.com/repos/huggingface/transformers/actions/runs/{run_id}/jobs?per_page=100'''
_A : List[Any] = requests.get(snake_case_ ).json()
_A : int = {}
try:
jobs.update({job["""name"""]: job["""html_url"""] for job in result["""jobs"""]} )
_A : Any = math.ceil((result["""total_count"""] - 100) / 100 )
for i in range(snake_case_ ):
_A : List[Any] = requests.get(url + f'''&page={i + 2}''' ).json()
jobs.update({job["""name"""]: job["""html_url"""] for job in result["""jobs"""]} )
return jobs
except Exception as e:
print("""Unknown error, could not fetch links.""",snake_case_ )
return {}
def lowerCAmelCase_ ( snake_case_ ):
_A : Any = {}
if os.path.exists(snake_case_ ):
_A : str = os.listdir(snake_case_ )
for file in files:
try:
with open(os.path.join(snake_case_,snake_case_ ),encoding="""utf-8""" ) as f:
_A : Dict = f.read()
except UnicodeDecodeError as e:
raise ValueError(f'''Could not open {os.path.join(snake_case_,snake_case_ )}.''' ) from e
return _artifact
def lowerCAmelCase_ ( ):
class lowercase :
def __init__( self , _a ) -> Tuple:
_A : List[str] = name
_A : Dict = []
def __str__( self ) -> str:
return self.name
def a__ ( self , _a ) -> str:
self.paths.append({"""name""": self.name, """path""": path} )
_A : Dict[str, Artifact] = {}
_A : Union[str, Any] = filter(os.path.isdir,os.listdir() )
for directory in directories:
_A : Any = directory
if artifact_name not in _available_artifacts:
_A : int = Artifact(snake_case_ )
_available_artifacts[artifact_name].add_path(snake_case_ )
return _available_artifacts
if __name__ == "__main__":
_snake_case = get_job_links()
_snake_case = retrieve_available_artifacts()
_snake_case = collections.OrderedDict(
[
("*.py", "API Examples"),
("*.md", "MD Examples"),
]
)
# This dict will contain all the information relative to each doc test category:
# - failed: list of failed tests
# - failures: dict in the format 'test': 'error_message'
_snake_case = {
v: {
"failed": [],
"failures": {},
}
for v in docs.values()
}
# Link to the GitHub Action job
_snake_case = github_actions_job_links.get("run_doctests")
_snake_case = available_artifacts["doc_tests_gpu_test_reports"].paths[0]
_snake_case = retrieve_artifact(artifact_path["name"])
if "stats" in artifact:
_snake_case , _snake_case , _snake_case = handle_test_results(artifact["stats"])
_snake_case = failed
_snake_case = success
_snake_case = time_spent[1:-1] + ", "
_snake_case = extract_first_line_failure(artifact["failures_short"])
for line in artifact["summary_short"].split("\n"):
if re.search("FAILED", line):
_snake_case = line.replace("FAILED ", "")
_snake_case = line.split()[0].replace("\n", "")
if "::" in line:
_snake_case , _snake_case = line.split("::")
else:
_snake_case , _snake_case = line, line
for file_regex in docs.keys():
if fnmatch(file_path, file_regex):
_snake_case = docs[file_regex]
doc_test_results[category]["failed"].append(test)
_snake_case = all_failures[test] if test in all_failures else "N/A"
_snake_case = failure
break
_snake_case = Message("🤗 Results of the doc tests.", doc_test_results)
message.post()
message.post_reply()
| 26 |
def lowerCAmelCase_ ( snake_case_ ):
if n_term == "":
return []
_A : list = []
for temp in range(int(snake_case_ ) ):
series.append(f'''1/{temp + 1}''' if series else """1""" )
return series
if __name__ == "__main__":
_snake_case = input("Enter the last number (nth term) of the Harmonic Series")
print("Formula of Harmonic Series => 1+1/2+1/3 ..... 1/n")
print(harmonic_series(nth_term))
| 26 | 1 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
if n == 0:
return 0
_A : Tuple = float("""-inf""" )
for i in range(1,n + 1 ):
_A : str = max(
snake_case_,prices[i - 1] + naive_cut_rod_recursive(n - i,snake_case_ ) )
return max_revue
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
_A : Dict = [float("""-inf""" ) for _ in range(n + 1 )]
return _top_down_cut_rod_recursive(snake_case_,snake_case_,snake_case_ )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
if max_rev[n] >= 0:
return max_rev[n]
elif n == 0:
return 0
else:
_A : List[str] = float("""-inf""" )
for i in range(1,n + 1 ):
_A : Optional[Any] = max(
snake_case_,prices[i - 1] + _top_down_cut_rod_recursive(n - i,snake_case_,snake_case_ ),)
_A : Tuple = max_revenue
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
# length(max_rev) = n + 1, to accommodate for the revenue obtainable from a rod of
# length 0.
_A : List[Any] = [float("""-inf""" ) for _ in range(n + 1 )]
_A : Any = 0
for i in range(1,n + 1 ):
_A : Optional[Any] = max_rev[i]
for j in range(1,i + 1 ):
_A : int = max(snake_case_,prices[j - 1] + max_rev[i - j] )
_A : int = max_revenue_i
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if n < 0:
_A : Optional[Any] = f'''n must be greater than or equal to 0. Got n = {n}'''
raise ValueError(snake_case_ )
if n > len(snake_case_ ):
_A : Any = (
"""Each integral piece of rod must have a corresponding price. """
f'''Got n = {n} but length of prices = {len(snake_case_ )}'''
)
raise ValueError(snake_case_ )
def lowerCAmelCase_ ( ):
_A : Tuple = [6, 10, 12, 15, 20, 23]
_A : List[Any] = len(snake_case_ )
# the best revenue comes from cutting the rod into 6 pieces, each
# of length 1 resulting in a revenue of 6 * 6 = 36.
_A : Any = 36
_A : List[Any] = top_down_cut_rod(snake_case_,snake_case_ )
_A : List[Any] = bottom_up_cut_rod(snake_case_,snake_case_ )
_A : Dict = naive_cut_rod_recursive(snake_case_,snake_case_ )
assert expected_max_revenue == max_rev_top_down
assert max_rev_top_down == max_rev_bottom_up
assert max_rev_bottom_up == max_rev_naive
if __name__ == "__main__":
main()
| 26 |
import importlib
import json
import os
from collections import OrderedDict
from typing import Dict, Optional, Union
# Build the list of all feature extractors
from ...configuration_utils import PretrainedConfig
from ...dynamic_module_utils import get_class_from_dynamic_module, resolve_trust_remote_code
from ...feature_extraction_utils import FeatureExtractionMixin
from ...utils import CONFIG_NAME, FEATURE_EXTRACTOR_NAME, get_file_from_repo, logging
from .auto_factory import _LazyAutoMapping
from .configuration_auto import (
CONFIG_MAPPING_NAMES,
AutoConfig,
model_type_to_module_name,
replace_list_option_in_docstrings,
)
_snake_case = logging.get_logger(__name__)
_snake_case = OrderedDict(
[
("audio-spectrogram-transformer", "ASTFeatureExtractor"),
("beit", "BeitFeatureExtractor"),
("chinese_clip", "ChineseCLIPFeatureExtractor"),
("clap", "ClapFeatureExtractor"),
("clip", "CLIPFeatureExtractor"),
("clipseg", "ViTFeatureExtractor"),
("conditional_detr", "ConditionalDetrFeatureExtractor"),
("convnext", "ConvNextFeatureExtractor"),
("cvt", "ConvNextFeatureExtractor"),
("data2vec-audio", "Wav2Vec2FeatureExtractor"),
("data2vec-vision", "BeitFeatureExtractor"),
("deformable_detr", "DeformableDetrFeatureExtractor"),
("deit", "DeiTFeatureExtractor"),
("detr", "DetrFeatureExtractor"),
("dinat", "ViTFeatureExtractor"),
("donut-swin", "DonutFeatureExtractor"),
("dpt", "DPTFeatureExtractor"),
("encodec", "EncodecFeatureExtractor"),
("flava", "FlavaFeatureExtractor"),
("glpn", "GLPNFeatureExtractor"),
("groupvit", "CLIPFeatureExtractor"),
("hubert", "Wav2Vec2FeatureExtractor"),
("imagegpt", "ImageGPTFeatureExtractor"),
("layoutlmv2", "LayoutLMv2FeatureExtractor"),
("layoutlmv3", "LayoutLMv3FeatureExtractor"),
("levit", "LevitFeatureExtractor"),
("maskformer", "MaskFormerFeatureExtractor"),
("mctct", "MCTCTFeatureExtractor"),
("mobilenet_v1", "MobileNetV1FeatureExtractor"),
("mobilenet_v2", "MobileNetV2FeatureExtractor"),
("mobilevit", "MobileViTFeatureExtractor"),
("nat", "ViTFeatureExtractor"),
("owlvit", "OwlViTFeatureExtractor"),
("perceiver", "PerceiverFeatureExtractor"),
("poolformer", "PoolFormerFeatureExtractor"),
("regnet", "ConvNextFeatureExtractor"),
("resnet", "ConvNextFeatureExtractor"),
("segformer", "SegformerFeatureExtractor"),
("sew", "Wav2Vec2FeatureExtractor"),
("sew-d", "Wav2Vec2FeatureExtractor"),
("speech_to_text", "Speech2TextFeatureExtractor"),
("speecht5", "SpeechT5FeatureExtractor"),
("swiftformer", "ViTFeatureExtractor"),
("swin", "ViTFeatureExtractor"),
("swinv2", "ViTFeatureExtractor"),
("table-transformer", "DetrFeatureExtractor"),
("timesformer", "VideoMAEFeatureExtractor"),
("tvlt", "TvltFeatureExtractor"),
("unispeech", "Wav2Vec2FeatureExtractor"),
("unispeech-sat", "Wav2Vec2FeatureExtractor"),
("van", "ConvNextFeatureExtractor"),
("videomae", "VideoMAEFeatureExtractor"),
("vilt", "ViltFeatureExtractor"),
("vit", "ViTFeatureExtractor"),
("vit_mae", "ViTFeatureExtractor"),
("vit_msn", "ViTFeatureExtractor"),
("wav2vec2", "Wav2Vec2FeatureExtractor"),
("wav2vec2-conformer", "Wav2Vec2FeatureExtractor"),
("wavlm", "Wav2Vec2FeatureExtractor"),
("whisper", "WhisperFeatureExtractor"),
("xclip", "CLIPFeatureExtractor"),
("yolos", "YolosFeatureExtractor"),
]
)
_snake_case = _LazyAutoMapping(CONFIG_MAPPING_NAMES, FEATURE_EXTRACTOR_MAPPING_NAMES)
def lowerCAmelCase_ ( snake_case_ ):
for module_name, extractors in FEATURE_EXTRACTOR_MAPPING_NAMES.items():
if class_name in extractors:
_A : List[str] = model_type_to_module_name(snake_case_ )
_A : List[Any] = importlib.import_module(f'''.{module_name}''',"""transformers.models""" )
try:
return getattr(snake_case_,snake_case_ )
except AttributeError:
continue
for _, extractor in FEATURE_EXTRACTOR_MAPPING._extra_content.items():
if getattr(snake_case_,"""__name__""",snake_case_ ) == class_name:
return extractor
# We did not fine the class, but maybe it's because a dep is missing. In that case, the class will be in the main
# init and we return the proper dummy to get an appropriate error message.
_A : List[Any] = importlib.import_module("""transformers""" )
if hasattr(snake_case_,snake_case_ ):
return getattr(snake_case_,snake_case_ )
return None
def lowerCAmelCase_ ( snake_case_,snake_case_ = None,snake_case_ = False,snake_case_ = False,snake_case_ = None,snake_case_ = None,snake_case_ = None,snake_case_ = False,**snake_case_,):
_A : Optional[int] = get_file_from_repo(
snake_case_,snake_case_,cache_dir=snake_case_,force_download=snake_case_,resume_download=snake_case_,proxies=snake_case_,use_auth_token=snake_case_,revision=snake_case_,local_files_only=snake_case_,)
if resolved_config_file is None:
logger.info(
"""Could not locate the feature extractor configuration file, will try to use the model config instead.""" )
return {}
with open(snake_case_,encoding="""utf-8""" ) as reader:
return json.load(snake_case_ )
class lowercase :
def __init__( self ) -> List[Any]:
raise EnvironmentError(
"""AutoFeatureExtractor is designed to be instantiated """
"""using the `AutoFeatureExtractor.from_pretrained(pretrained_model_name_or_path)` method.""" )
@classmethod
@replace_list_option_in_docstrings(_a )
def a__ ( cls , _a , **_a ) -> Any:
_A : Tuple = kwargs.pop("""config""" , _a )
_A : Tuple = kwargs.pop("""trust_remote_code""" , _a )
_A : List[Any] = True
_A , _A : Tuple = FeatureExtractionMixin.get_feature_extractor_dict(_a , **_a )
_A : Tuple = config_dict.get("""feature_extractor_type""" , _a )
_A : int = None
if "AutoFeatureExtractor" in config_dict.get("""auto_map""" , {} ):
_A : Optional[int] = config_dict["""auto_map"""]["""AutoFeatureExtractor"""]
# If we don't find the feature extractor class in the feature extractor config, let's try the model config.
if feature_extractor_class is None and feature_extractor_auto_map is None:
if not isinstance(_a , _a ):
_A : int = AutoConfig.from_pretrained(_a , **_a )
# It could be in `config.feature_extractor_type``
_A : Optional[int] = getattr(_a , """feature_extractor_type""" , _a )
if hasattr(_a , """auto_map""" ) and "AutoFeatureExtractor" in config.auto_map:
_A : Tuple = config.auto_map["""AutoFeatureExtractor"""]
if feature_extractor_class is not None:
_A : Optional[Any] = feature_extractor_class_from_name(_a )
_A : List[Any] = feature_extractor_auto_map is not None
_A : Union[str, Any] = feature_extractor_class is not None or type(_a ) in FEATURE_EXTRACTOR_MAPPING
_A : Optional[int] = resolve_trust_remote_code(
_a , _a , _a , _a )
if has_remote_code and trust_remote_code:
_A : Dict = get_class_from_dynamic_module(
_a , _a , **_a )
_A : str = kwargs.pop("""code_revision""" , _a )
if os.path.isdir(_a ):
feature_extractor_class.register_for_auto_class()
return feature_extractor_class.from_dict(_a , **_a )
elif feature_extractor_class is not None:
return feature_extractor_class.from_dict(_a , **_a )
# Last try: we use the FEATURE_EXTRACTOR_MAPPING.
elif type(_a ) in FEATURE_EXTRACTOR_MAPPING:
_A : Dict = FEATURE_EXTRACTOR_MAPPING[type(_a )]
return feature_extractor_class.from_dict(_a , **_a )
raise ValueError(
F'''Unrecognized feature extractor in {pretrained_model_name_or_path}. Should have a '''
F'''`feature_extractor_type` key in its {FEATURE_EXTRACTOR_NAME} of {CONFIG_NAME}, or one of the following '''
F'''`model_type` keys in its {CONFIG_NAME}: {", ".join(c for c in FEATURE_EXTRACTOR_MAPPING_NAMES.keys() )}''' )
@staticmethod
def a__ ( _a , _a ) -> Optional[int]:
FEATURE_EXTRACTOR_MAPPING.register(_a , _a )
| 26 | 1 |
# Lint as: python3
# pylint: enable=line-too-long
# pylint: disable=g-import-not-at-top,g-bad-import-order,wrong-import-position
_snake_case = "2.13.1"
import platform
import pyarrow
from packaging import version
if version.parse(platform.python_version()) < version.parse("3.7"):
raise ImportWarning(
"To use `datasets`, Python>=3.7 is required, and the current version of Python doesn't match this condition."
)
if version.parse(pyarrow.__version__).major < 8:
raise ImportWarning(
"To use `datasets`, the module `pyarrow>=8.0.0` is required, and the current version of `pyarrow` doesn't match this condition.\n"
"If you are running this in a Google Colab, you should probably just restart the runtime to use the right version of `pyarrow`."
)
del platform
del pyarrow
del version
from .arrow_dataset import Dataset
from .arrow_reader import ReadInstruction
from .builder import ArrowBasedBuilder, BeamBasedBuilder, BuilderConfig, DatasetBuilder, GeneratorBasedBuilder
from .combine import concatenate_datasets, interleave_datasets
from .dataset_dict import DatasetDict, IterableDatasetDict
from .download import *
from .features import *
from .fingerprint import disable_caching, enable_caching, is_caching_enabled, set_caching_enabled
from .info import DatasetInfo, MetricInfo
from .inspect import (
get_dataset_config_info,
get_dataset_config_names,
get_dataset_infos,
get_dataset_split_names,
inspect_dataset,
inspect_metric,
list_datasets,
list_metrics,
)
from .iterable_dataset import IterableDataset
from .load import load_dataset, load_dataset_builder, load_from_disk, load_metric
from .metric import Metric
from .splits import (
NamedSplit,
NamedSplitAll,
Split,
SplitBase,
SplitDict,
SplitGenerator,
SplitInfo,
SubSplitInfo,
percent,
)
from .tasks import *
from .utils import *
from .utils import logging
# deprecated modules
from datasets import arrow_dataset as _arrow_dataset # isort:skip
from datasets import utils as _utils # isort:skip
from datasets.utils import download_manager as _deprecated_download_manager # isort:skip
_snake_case = concatenate_datasets
_snake_case = DownloadConfig
_snake_case = DownloadManager
_snake_case = DownloadMode
_snake_case = DownloadConfig
_snake_case = DownloadMode
_snake_case = DownloadManager
del _arrow_dataset, _utils, _deprecated_download_manager
| 26 |
import unittest
import numpy as np
from transformers.testing_utils import is_flaky, require_torch, require_vision
from transformers.utils import is_torch_available, is_vision_available
from ...test_image_processing_common import ImageProcessingSavingTestMixin, prepare_image_inputs
if is_torch_available():
import torch
if is_vision_available():
from PIL import Image
from transformers import DonutImageProcessor
class lowercase ( unittest.TestCase ):
def __init__( self , _a , _a=7 , _a=3 , _a=18 , _a=30 , _a=400 , _a=True , _a=None , _a=True , _a=False , _a=True , _a=True , _a=[0.5, 0.5, 0.5] , _a=[0.5, 0.5, 0.5] , ) -> Dict:
_A : str = parent
_A : int = batch_size
_A : Optional[int] = num_channels
_A : List[Any] = image_size
_A : int = min_resolution
_A : Optional[int] = max_resolution
_A : Any = do_resize
_A : List[str] = size if size is not None else {"""height""": 18, """width""": 20}
_A : Optional[int] = do_thumbnail
_A : str = do_align_axis
_A : List[Any] = do_pad
_A : Optional[Any] = do_normalize
_A : Tuple = image_mean
_A : List[str] = image_std
def a__ ( self ) -> Optional[int]:
return {
"do_resize": self.do_resize,
"size": self.size,
"do_thumbnail": self.do_thumbnail,
"do_align_long_axis": self.do_align_axis,
"do_pad": self.do_pad,
"do_normalize": self.do_normalize,
"image_mean": self.image_mean,
"image_std": self.image_std,
}
@require_torch
@require_vision
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = DonutImageProcessor if is_vision_available() else None
def a__ ( self ) -> Optional[int]:
_A : List[str] = DonutImageProcessingTester(self )
@property
def a__ ( self ) -> List[Any]:
return self.image_processor_tester.prepare_image_processor_dict()
def a__ ( self ) -> Optional[Any]:
_A : Union[str, Any] = self.image_processing_class(**self.image_processor_dict )
self.assertTrue(hasattr(_a , """do_resize""" ) )
self.assertTrue(hasattr(_a , """size""" ) )
self.assertTrue(hasattr(_a , """do_thumbnail""" ) )
self.assertTrue(hasattr(_a , """do_align_long_axis""" ) )
self.assertTrue(hasattr(_a , """do_pad""" ) )
self.assertTrue(hasattr(_a , """do_normalize""" ) )
self.assertTrue(hasattr(_a , """image_mean""" ) )
self.assertTrue(hasattr(_a , """image_std""" ) )
def a__ ( self ) -> List[Any]:
_A : Union[str, Any] = self.image_processing_class.from_dict(self.image_processor_dict )
self.assertEqual(image_processor.size , {"""height""": 18, """width""": 20} )
_A : Optional[int] = self.image_processing_class.from_dict(self.image_processor_dict , size=42 )
self.assertEqual(image_processor.size , {"""height""": 42, """width""": 42} )
# Previous config had dimensions in (width, height) order
_A : List[str] = self.image_processing_class.from_dict(self.image_processor_dict , size=(42, 84) )
self.assertEqual(image_processor.size , {"""height""": 84, """width""": 42} )
def a__ ( self ) -> Union[str, Any]:
pass
@is_flaky()
def a__ ( self ) -> Optional[int]:
# Initialize image_processing
_A : List[Any] = self.image_processing_class(**self.image_processor_dict )
# create random PIL images
_A : Dict = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a )
for image in image_inputs:
self.assertIsInstance(_a , Image.Image )
# Test not batched input
_A : Union[str, Any] = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : Any = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
@is_flaky()
def a__ ( self ) -> Dict:
# Initialize image_processing
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random numpy tensors
_A : Tuple = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , numpify=_a )
for image in image_inputs:
self.assertIsInstance(_a , np.ndarray )
# Test not batched input
_A : int = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : List[str] = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
@is_flaky()
def a__ ( self ) -> Optional[int]:
# Initialize image_processing
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random PyTorch tensors
_A : Optional[int] = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , torchify=_a )
for image in image_inputs:
self.assertIsInstance(_a , torch.Tensor )
# Test not batched input
_A : Any = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : str = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
| 26 | 1 |
import warnings
from diffusers import StableDiffusionImgaImgPipeline # noqa F401
warnings.warn(
"The `image_to_image.py` script is outdated. Please use directly `from diffusers import"
" StableDiffusionImg2ImgPipeline` instead."
)
| 26 |
from __future__ import annotations
import numpy as np
def lowerCAmelCase_ ( snake_case_ ):
return np.maximum(0,snake_case_ )
if __name__ == "__main__":
print(np.array(relu([-1, 0, 5]))) # --> [0, 0, 5]
| 26 | 1 |
from __future__ import annotations
import unittest
from transformers import AutoTokenizer, PegasusConfig, is_tf_available
from transformers.testing_utils import require_sentencepiece, require_tf, require_tokenizers, slow
from transformers.utils import cached_property
from ...test_configuration_common import ConfigTester
from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_tf_available():
import tensorflow as tf
from transformers import TFAutoModelForSeqaSeqLM, TFPegasusForConditionalGeneration, TFPegasusModel
@require_tf
class lowercase :
_a = PegasusConfig
_a = {}
_a = "gelu"
def __init__( self , _a , _a=13 , _a=7 , _a=True , _a=False , _a=99 , _a=32 , _a=2 , _a=4 , _a=37 , _a=0.1 , _a=0.1 , _a=40 , _a=2 , _a=1 , _a=0 , ) -> List[str]:
_A : List[Any] = parent
_A : Optional[Any] = batch_size
_A : Tuple = seq_length
_A : int = is_training
_A : List[Any] = use_labels
_A : int = vocab_size
_A : Dict = hidden_size
_A : Dict = num_hidden_layers
_A : int = num_attention_heads
_A : int = intermediate_size
_A : Dict = hidden_dropout_prob
_A : Dict = attention_probs_dropout_prob
_A : Union[str, Any] = max_position_embeddings
_A : int = eos_token_id
_A : Dict = pad_token_id
_A : List[str] = bos_token_id
def a__ ( self ) -> str:
_A : str = ids_tensor([self.batch_size, self.seq_length - 1] , self.vocab_size )
_A : Optional[Any] = tf.expand_dims(tf.constant([self.eos_token_id] * self.batch_size ) , 1 )
_A : Any = tf.concat([input_ids, eos_tensor] , axis=1 )
_A : Optional[int] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
_A : Optional[Any] = self.config_cls(
vocab_size=self.vocab_size , d_model=self.hidden_size , encoder_layers=self.num_hidden_layers , decoder_layers=self.num_hidden_layers , encoder_attention_heads=self.num_attention_heads , decoder_attention_heads=self.num_attention_heads , encoder_ffn_dim=self.intermediate_size , decoder_ffn_dim=self.intermediate_size , dropout=self.hidden_dropout_prob , attention_dropout=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , eos_token_ids=[2] , bos_token_id=self.bos_token_id , pad_token_id=self.pad_token_id , decoder_start_token_id=self.pad_token_id , **self.config_updates , )
_A : List[Any] = prepare_pegasus_inputs_dict(_a , _a , _a )
return config, inputs_dict
def a__ ( self , _a , _a ) -> Optional[Any]:
_A : Any = TFPegasusModel(config=_a ).get_decoder()
_A : Union[str, Any] = inputs_dict["""input_ids"""]
_A : Union[str, Any] = input_ids[:1, :]
_A : List[str] = inputs_dict["""attention_mask"""][:1, :]
_A : Any = inputs_dict["""head_mask"""]
_A : Any = 1
# first forward pass
_A : int = model(_a , attention_mask=_a , head_mask=_a , use_cache=_a )
_A , _A : Union[str, Any] = outputs.to_tuple()
# create hypothetical next token and extent to next_input_ids
_A : List[str] = ids_tensor((self.batch_size, 3) , config.vocab_size )
_A : Optional[Any] = tf.cast(ids_tensor((self.batch_size, 3) , 2 ) , tf.inta )
# append to next input_ids and
_A : Dict = tf.concat([input_ids, next_tokens] , axis=-1 )
_A : int = tf.concat([attention_mask, next_attn_mask] , axis=-1 )
_A : Any = model(_a , attention_mask=_a )[0]
_A : List[Any] = model(_a , attention_mask=_a , past_key_values=_a )[0]
self.parent.assertEqual(next_tokens.shape[1] , output_from_past.shape[1] )
# select random slice
_A : List[Any] = int(ids_tensor((1,) , output_from_past.shape[-1] ) )
_A : Optional[int] = output_from_no_past[:, -3:, random_slice_idx]
_A : int = output_from_past[:, :, random_slice_idx]
# test that outputs are equal for slice
tf.debugging.assert_near(_a , _a , rtol=1e-3 )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_=None,snake_case_=None,snake_case_=None,snake_case_=None,snake_case_=None,):
if attention_mask is None:
_A : List[str] = tf.cast(tf.math.not_equal(snake_case_,config.pad_token_id ),tf.inta )
if decoder_attention_mask is None:
_A : int = tf.concat(
[
tf.ones(decoder_input_ids[:, :1].shape,dtype=tf.inta ),
tf.cast(tf.math.not_equal(decoder_input_ids[:, 1:],config.pad_token_id ),tf.inta ),
],axis=-1,)
if head_mask is None:
_A : List[Any] = tf.ones((config.encoder_layers, config.encoder_attention_heads) )
if decoder_head_mask is None:
_A : Dict = tf.ones((config.decoder_layers, config.decoder_attention_heads) )
if cross_attn_head_mask is None:
_A : Optional[int] = tf.ones((config.decoder_layers, config.decoder_attention_heads) )
return {
"input_ids": input_ids,
"decoder_input_ids": decoder_input_ids,
"attention_mask": attention_mask,
"decoder_attention_mask": decoder_attention_mask,
"head_mask": head_mask,
"decoder_head_mask": decoder_head_mask,
"cross_attn_head_mask": cross_attn_head_mask,
}
@require_tf
class lowercase ( UpperCamelCase__,UpperCamelCase__,unittest.TestCase ):
_a = (TFPegasusForConditionalGeneration, TFPegasusModel) if is_tf_available() else ()
_a = (TFPegasusForConditionalGeneration,) if is_tf_available() else ()
_a = (
{
"conversational": TFPegasusForConditionalGeneration,
"feature-extraction": TFPegasusModel,
"summarization": TFPegasusForConditionalGeneration,
"text2text-generation": TFPegasusForConditionalGeneration,
"translation": TFPegasusForConditionalGeneration,
}
if is_tf_available()
else {}
)
_a = True
_a = False
_a = False
def a__ ( self ) -> Optional[Any]:
_A : Tuple = TFPegasusModelTester(self )
_A : List[Any] = ConfigTester(self , config_class=_a )
def a__ ( self ) -> Optional[int]:
self.config_tester.run_common_tests()
def a__ ( self ) -> Any:
_A : Union[str, Any] = self.model_tester.prepare_config_and_inputs_for_common()
self.model_tester.check_decoder_model_past_large_inputs(*_a )
@require_sentencepiece
@require_tokenizers
@require_tf
class lowercase ( unittest.TestCase ):
_a = [
" PG&E stated it scheduled the blackouts in response to forecasts for high winds amid dry conditions. The aim is to reduce the risk of wildfires. Nearly 800 thousand customers were scheduled to be affected by the shutoffs which were expected to last through at least midday tomorrow.",
" The London trio are up for best UK act and best album, as well as getting two nominations in the best song category.\"We got told like this morning 'Oh I think you're nominated'\", said Dappy.\"And I was like 'Oh yeah, which one?' And now we've got nominated for four awards. I mean, wow!\"Bandmate Fazer added: \"We thought it's best of us to come down and mingle with everyone and say hello to the cameras. And now we find we've got four nominations.\"The band have two shots at the best song prize, getting the nod for their Tynchy Stryder collaboration Number One, and single Strong Again.Their album Uncle B will also go up against records by the likes of Beyonce and Kanye West.N-Dubz picked up the best newcomer Mobo in 2007, but female member Tulisa said they wouldn't be too disappointed if they didn't win this time around.\"At the end of the day we're grateful to be where we are in our careers.\"If it don't happen then it don't happen - live to fight another day and keep on making albums and hits for the fans.\"Dappy also revealed they could be performing live several times on the night.The group will be doing Number One and also a possible rendition of the War Child single, I Got Soul.The charity song is a re-working of The Killers' All These Things That I've Done and is set to feature artists like Chipmunk, Ironik and Pixie Lott.This year's Mobos will be held outside of London for the first time, in Glasgow on 30 September.N-Dubz said they were looking forward to performing for their Scottish fans and boasted about their recent shows north of the border.\"We just done Edinburgh the other day,\" said Dappy.\"We smashed up an N-Dubz show over there. We done Aberdeen about three or four months ago - we smashed up that show over there! Everywhere we go we smash it up!\" ",
]
_a = [
"California's largest electricity provider has cut power to hundreds of thousands of customers in an effort to"
" reduce the risk of wildfires.",
"N-Dubz have revealed they\'re \"grateful\" to have been nominated for four Mobo Awards.",
] # differs slightly from pytorch, likely due to numerical differences in linear layers
_a = "google/pegasus-xsum"
@cached_property
def a__ ( self ) -> List[Any]:
return AutoTokenizer.from_pretrained(self.model_name )
@cached_property
def a__ ( self ) -> List[Any]:
_A : str = TFAutoModelForSeqaSeqLM.from_pretrained(self.model_name )
return model
def a__ ( self , **_a ) -> Tuple:
_A : Dict = self.translate_src_text(**_a )
assert self.expected_text == generated_words
def a__ ( self , **_a ) -> Any:
_A : str = self.tokenizer(self.src_text , **_a , padding=_a , return_tensors="""tf""" )
_A : Tuple = self.model.generate(
model_inputs.input_ids , attention_mask=model_inputs.attention_mask , num_beams=2 , use_cache=_a , )
_A : List[str] = self.tokenizer.batch_decode(generated_ids.numpy() , skip_special_tokens=_a )
return generated_words
@slow
def a__ ( self ) -> Optional[int]:
self._assert_generated_batch_equal_expected()
| 26 |
import argparse
import shutil
import time
from json import JSONDecodeError
from logging import getLogger
from pathlib import Path
from typing import Dict, List
import torch
from torch.utils.data import DataLoader
from tqdm import tqdm
from transformers import AutoModelForSeqaSeqLM, AutoTokenizer
from utils import (
SeqaSeqDataset,
calculate_bleu,
calculate_rouge,
chunks,
lmap,
load_json,
parse_numeric_n_bool_cl_kwargs,
save_json,
use_task_specific_params,
write_txt_file,
)
_snake_case = getLogger(__name__)
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ = 8,snake_case_ = 1024,snake_case_="val",snake_case_=None,snake_case_=False,snake_case_="summarization",snake_case_=None,snake_case_=1,snake_case_ = None,snake_case_="",**snake_case_,):
_A : Dict = str(snake_case_ )
assert local_rank is not None
torch.distributed.init_process_group(backend="""nccl""",rank=snake_case_ )
_A : Tuple = Path(snake_case_ )
_A : List[Any] = save_dir.joinpath(f'''rank_{local_rank}_output.json''' )
torch.cuda.set_device(snake_case_ )
_A : Union[str, Any] = AutoModelForSeqaSeqLM.from_pretrained(snake_case_ ).cuda()
if fpaa:
_A : Any = model.half()
# determine if we need to increase num_beams
use_task_specific_params(snake_case_,snake_case_ ) # update config with task specific params
_A : str = generate_kwargs.pop("""num_beams""",model.config.num_beams ) # AttributeError risk?
if num_return_sequences > num_beams:
_A : int = num_return_sequences
_A : Optional[Any] = AutoTokenizer.from_pretrained(snake_case_ )
logger.info(f'''Inferred tokenizer type: {tokenizer.__class__}''' ) # if this is wrong, check config.model_type.
if max_source_length is None:
_A : Optional[int] = tokenizer.model_max_length
if prefix is None:
_A : Tuple = prefix or getattr(model.config,"""prefix""","""""" ) or """"""
_A : Optional[int] = SeqaSeqDataset(
snake_case_,snake_case_,snake_case_,max_target_length=1024,type_path=snake_case_,n_obs=snake_case_,prefix=snake_case_,**snake_case_,)
# I set shuffle=True for a more accurate progress bar.
# If all the longest samples are first, the prog bar estimate is too high at the beginning.
_A : Optional[int] = ds.make_sortish_sampler(snake_case_,distributed=snake_case_,add_extra_examples=snake_case_,shuffle=snake_case_ )
_A : Dict = DataLoader(snake_case_,sampler=snake_case_,batch_size=snake_case_,collate_fn=ds.collate_fn )
_A : Optional[Any] = []
for batch in tqdm(snake_case_ ):
_A : Tuple = model.generate(
input_ids=batch["""input_ids"""].to(model.device ),attention_mask=batch["""attention_mask"""].to(model.device ),num_return_sequences=snake_case_,num_beams=snake_case_,**snake_case_,)
_A : Any = tokenizer.batch_decode(snake_case_,skip_special_tokens=snake_case_,clean_up_tokenization_spaces=snake_case_ )
_A : Dict = batch["""ids"""]
if num_return_sequences > 1:
_A : Any = chunks(snake_case_,snake_case_ ) # batch size chunks, each of size num_return_seq
for i, pred in enumerate(snake_case_ ):
results.append({"""pred""": pred, """id""": ids[i].item()} )
save_json(snake_case_,snake_case_ )
return results, sampler.num_replicas
def lowerCAmelCase_ ( ):
_A : Tuple = argparse.ArgumentParser(
epilog="""Unspecified args like --num_beams=2 --decoder_start_token_id=4 are passed to model.generate""" )
parser.add_argument("""--data_dir""",type=snake_case_,help="""like cnn_dm/test.source""" )
parser.add_argument(
"""--model_name""",type=snake_case_,help="""like facebook/bart-large-cnn,t5-base, etc.""",default="""sshleifer/distilbart-xsum-12-3""",)
parser.add_argument("""--save_dir""",type=snake_case_,help="""where to save""",default="""tmp_gen""" )
parser.add_argument("""--max_source_length""",type=snake_case_,default=snake_case_ )
parser.add_argument(
"""--type_path""",type=snake_case_,default="""test""",help="""which subset to evaluate typically train/val/test""" )
parser.add_argument("""--task""",type=snake_case_,default="""summarization""",help="""used for task_specific_params + metrics""" )
parser.add_argument("""--bs""",type=snake_case_,default=8,required=snake_case_,help="""batch size""" )
parser.add_argument(
"""--local_rank""",type=snake_case_,default=-1,required=snake_case_,help="""should be passed by distributed.launch""" )
parser.add_argument(
"""--n_obs""",type=snake_case_,default=snake_case_,required=snake_case_,help="""How many observations. Defaults to all.""" )
parser.add_argument(
"""--num_return_sequences""",type=snake_case_,default=1,required=snake_case_,help="""How many sequences to return""" )
parser.add_argument(
"""--sync_timeout""",type=snake_case_,default=600,required=snake_case_,help="""How long should master process wait for other processes to finish.""",)
parser.add_argument("""--src_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument("""--tgt_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument(
"""--prefix""",type=snake_case_,required=snake_case_,default=snake_case_,help="""will be added to the begininng of src examples""" )
parser.add_argument("""--fp16""",action="""store_true""" )
parser.add_argument("""--debug""",action="""store_true""" )
_A : Union[str, Any] = time.time()
_A , _A : List[str] = parser.parse_known_args()
_A : List[str] = parse_numeric_n_bool_cl_kwargs(snake_case_ )
if generate_kwargs and args.local_rank <= 0:
print(f'''parsed the following generate kwargs: {generate_kwargs}''' )
_A : Dict = Path(args.save_dir + """_tmp""" )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ ) # this handles locking.
_A : int = list(json_save_dir.glob("""rank_*.json""" ) )
if intermediate_files:
raise ValueError(f'''Found files at {json_save_dir} please move or remove them.''' )
# In theory, a node could finish and save before another node hits this. If this happens, we can address later.
_A : Any = {}
if args.src_lang is not None:
_A : int = args.src_lang
if args.tgt_lang is not None:
_A : Dict = args.tgt_lang
Path(args.save_dir ).mkdir(exist_ok=snake_case_ )
_A , _A : str = eval_data_dir(
args.data_dir,snake_case_,args.model_name,type_path=args.type_path,bs=args.bs,fpaa=args.fpaa,task=args.task,local_rank=args.local_rank,n_obs=args.n_obs,max_source_length=args.max_source_length,num_return_sequences=args.num_return_sequences,prefix=args.prefix,dataset_kwargs=snake_case_,**snake_case_,)
if args.local_rank <= 0:
_A : List[Any] = Path(args.save_dir )
save_dir.mkdir(exist_ok=snake_case_ )
_A : Tuple = gather_results_from_each_node(snake_case_,snake_case_,args.sync_timeout )
_A : Optional[int] = combine_partial_results(snake_case_ )
if args.num_return_sequences > 1:
_A : Optional[Any] = save_dir.joinpath("""pseudolabel_results.json""" )
print(f'''Saving aggregated results at {save_path}, intermediate in {json_save_dir}/''' )
save_json(snake_case_,snake_case_ )
return
_A : List[str] = Path(args.data_dir ).joinpath(args.type_path + """.target""" )
with open(snake_case_ ) as f:
_A : int = [x.rstrip() for x in f.readlines()][: len(snake_case_ )]
# Calculate metrics, save metrics, and save _generations.txt
_A : Dict = """translation""" in args.task
_A : Optional[Any] = calculate_bleu if calc_bleu else calculate_rouge
_A : Tuple = """bleu""" if calc_bleu else """rouge"""
_A : Dict = score_fn(snake_case_,snake_case_ )
_A : List[Any] = len(snake_case_ )
_A : Optional[int] = time.time() - start_time
_A : Dict = round(runtime / metrics["""n_obs"""],4 )
_A : Dict = num_replicas
# TODO(@stas00): add whatever metadata to metrics
_A : Any = save_dir.joinpath(f'''{args.type_path}_{metric_name}.json''' )
save_json(snake_case_,snake_case_,indent=snake_case_ )
print(snake_case_ )
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}_generations.txt''' ) )
if args.debug:
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}.target''' ) )
else:
shutil.rmtree(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
_A : Dict = []
for partial_result in partial_results:
records.extend(snake_case_ )
_A : Optional[Any] = sorted(snake_case_,key=lambda snake_case_ : x["id"] )
_A : List[str] = [x["""pred"""] for x in records]
return preds
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
# WAIT FOR lots of .json files
_A : Optional[Any] = time.time()
logger.info("""waiting for all nodes to finish""" )
_A : List[str] = None
while (time.time() - start_wait) < timeout:
_A : str = list(save_dir.glob("""rank_*.json""" ) )
if len(snake_case_ ) < num_replicas:
continue
try:
# make sure all json files are fully saved
_A : List[str] = lmap(snake_case_,snake_case_ )
return json_data
except JSONDecodeError:
continue
else:
raise TimeoutError("""Rank 0 gave up on waiting for other processes""" )
# Unreachable
if __name__ == "__main__":
# Usage for MT:
run_generate()
| 26 | 1 |
import gc
import unittest
import numpy as np
import torch
from torch.backends.cuda import sdp_kernel
from diffusers import (
CMStochasticIterativeScheduler,
ConsistencyModelPipeline,
UNetaDModel,
)
from diffusers.utils import randn_tensor, slow, torch_device
from diffusers.utils.testing_utils import enable_full_determinism, require_torch_a, require_torch_gpu
from ..pipeline_params import UNCONDITIONAL_IMAGE_GENERATION_BATCH_PARAMS, UNCONDITIONAL_IMAGE_GENERATION_PARAMS
from ..test_pipelines_common import PipelineTesterMixin
enable_full_determinism()
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = ConsistencyModelPipeline
_a = UNCONDITIONAL_IMAGE_GENERATION_PARAMS
_a = UNCONDITIONAL_IMAGE_GENERATION_BATCH_PARAMS
# Override required_optional_params to remove num_images_per_prompt
_a = frozenset(
[
"num_inference_steps",
"generator",
"latents",
"output_type",
"return_dict",
"callback",
"callback_steps",
] )
@property
def a__ ( self ) -> Dict:
_A : List[str] = UNetaDModel.from_pretrained(
"""diffusers/consistency-models-test""" , subfolder="""test_unet""" , )
return unet
@property
def a__ ( self ) -> List[str]:
_A : Optional[Any] = UNetaDModel.from_pretrained(
"""diffusers/consistency-models-test""" , subfolder="""test_unet_class_cond""" , )
return unet
def a__ ( self , _a=False ) -> List[Any]:
if class_cond:
_A : List[str] = self.dummy_cond_unet
else:
_A : List[str] = self.dummy_uncond_unet
# Default to CM multistep sampler
_A : int = CMStochasticIterativeScheduler(
num_train_timesteps=40 , sigma_min=0.002 , sigma_max=80.0 , )
_A : List[Any] = {
"""unet""": unet,
"""scheduler""": scheduler,
}
return components
def a__ ( self , _a , _a=0 ) -> Any:
if str(_a ).startswith("""mps""" ):
_A : int = torch.manual_seed(_a )
else:
_A : Union[str, Any] = torch.Generator(device=_a ).manual_seed(_a )
_A : List[str] = {
"""batch_size""": 1,
"""num_inference_steps""": None,
"""timesteps""": [22, 0],
"""generator""": generator,
"""output_type""": """np""",
}
return inputs
def a__ ( self ) -> Any:
_A : Any = """cpu""" # ensure determinism for the device-dependent torch.Generator
_A : Union[str, Any] = self.get_dummy_components()
_A : Any = ConsistencyModelPipeline(**_a )
_A : Any = pipe.to(_a )
pipe.set_progress_bar_config(disable=_a )
_A : Optional[int] = self.get_dummy_inputs(_a )
_A : Any = pipe(**_a ).images
assert image.shape == (1, 32, 32, 3)
_A : Union[str, Any] = image[0, -3:, -3:, -1]
_A : Dict = np.array([0.3572, 0.6273, 0.4031, 0.3961, 0.4321, 0.5730, 0.5266, 0.4780, 0.5004] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-3
def a__ ( self ) -> Dict:
_A : List[Any] = """cpu""" # ensure determinism for the device-dependent torch.Generator
_A : Any = self.get_dummy_components(class_cond=_a )
_A : str = ConsistencyModelPipeline(**_a )
_A : List[str] = pipe.to(_a )
pipe.set_progress_bar_config(disable=_a )
_A : List[str] = self.get_dummy_inputs(_a )
_A : Tuple = 0
_A : int = pipe(**_a ).images
assert image.shape == (1, 32, 32, 3)
_A : Union[str, Any] = image[0, -3:, -3:, -1]
_A : List[Any] = np.array([0.3572, 0.6273, 0.4031, 0.3961, 0.4321, 0.5730, 0.5266, 0.4780, 0.5004] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-3
def a__ ( self ) -> Any:
_A : Optional[int] = """cpu""" # ensure determinism for the device-dependent torch.Generator
_A : Tuple = self.get_dummy_components()
_A : Dict = ConsistencyModelPipeline(**_a )
_A : Optional[Any] = pipe.to(_a )
pipe.set_progress_bar_config(disable=_a )
_A : Union[str, Any] = self.get_dummy_inputs(_a )
_A : Optional[Any] = 1
_A : Optional[Any] = None
_A : Dict = pipe(**_a ).images
assert image.shape == (1, 32, 32, 3)
_A : Optional[int] = image[0, -3:, -3:, -1]
_A : Optional[Any] = np.array([0.5004, 0.5004, 0.4994, 0.5008, 0.4976, 0.5018, 0.4990, 0.4982, 0.4987] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-3
def a__ ( self ) -> List[str]:
_A : Tuple = """cpu""" # ensure determinism for the device-dependent torch.Generator
_A : Any = self.get_dummy_components(class_cond=_a )
_A : str = ConsistencyModelPipeline(**_a )
_A : List[str] = pipe.to(_a )
pipe.set_progress_bar_config(disable=_a )
_A : Optional[int] = self.get_dummy_inputs(_a )
_A : Optional[int] = 1
_A : Any = None
_A : str = 0
_A : Dict = pipe(**_a ).images
assert image.shape == (1, 32, 32, 3)
_A : str = image[0, -3:, -3:, -1]
_A : str = np.array([0.5004, 0.5004, 0.4994, 0.5008, 0.4976, 0.5018, 0.4990, 0.4982, 0.4987] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-3
@slow
@require_torch_gpu
class lowercase ( unittest.TestCase ):
def a__ ( self ) -> Tuple:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self , _a=0 , _a=False , _a="cpu" , _a=torch.floataa , _a=(1, 3, 64, 64) ) -> Optional[int]:
_A : Optional[int] = torch.manual_seed(_a )
_A : List[Any] = {
"""num_inference_steps""": None,
"""timesteps""": [22, 0],
"""class_labels""": 0,
"""generator""": generator,
"""output_type""": """np""",
}
if get_fixed_latents:
_A : Tuple = self.get_fixed_latents(seed=_a , device=_a , dtype=_a , shape=_a )
_A : Any = latents
return inputs
def a__ ( self , _a=0 , _a="cpu" , _a=torch.floataa , _a=(1, 3, 64, 64) ) -> Optional[Any]:
if type(_a ) == str:
_A : int = torch.device(_a )
_A : Optional[int] = torch.Generator(device=_a ).manual_seed(_a )
_A : Tuple = randn_tensor(_a , generator=_a , device=_a , dtype=_a )
return latents
def a__ ( self ) -> Any:
_A : List[str] = UNetaDModel.from_pretrained("""diffusers/consistency_models""" , subfolder="""diffusers_cd_imagenet64_l2""" )
_A : Optional[Any] = CMStochasticIterativeScheduler(
num_train_timesteps=40 , sigma_min=0.002 , sigma_max=80.0 , )
_A : Dict = ConsistencyModelPipeline(unet=_a , scheduler=_a )
pipe.to(torch_device=_a )
pipe.set_progress_bar_config(disable=_a )
_A : List[Any] = self.get_inputs()
_A : Optional[Any] = pipe(**_a ).images
assert image.shape == (1, 64, 64, 3)
_A : int = image[0, -3:, -3:, -1]
_A : Dict = np.array([0.0888, 0.0881, 0.0666, 0.0479, 0.0292, 0.0195, 0.0201, 0.0163, 0.0254] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 2e-2
def a__ ( self ) -> Optional[int]:
_A : Any = UNetaDModel.from_pretrained("""diffusers/consistency_models""" , subfolder="""diffusers_cd_imagenet64_l2""" )
_A : Tuple = CMStochasticIterativeScheduler(
num_train_timesteps=40 , sigma_min=0.002 , sigma_max=80.0 , )
_A : Dict = ConsistencyModelPipeline(unet=_a , scheduler=_a )
pipe.to(torch_device=_a )
pipe.set_progress_bar_config(disable=_a )
_A : Union[str, Any] = self.get_inputs()
_A : Optional[Any] = 1
_A : str = None
_A : int = pipe(**_a ).images
assert image.shape == (1, 64, 64, 3)
_A : Optional[int] = image[0, -3:, -3:, -1]
_A : Tuple = np.array([0.0340, 0.0152, 0.0063, 0.0267, 0.0221, 0.0107, 0.0416, 0.0186, 0.0217] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 2e-2
@require_torch_a
def a__ ( self ) -> Optional[Any]:
_A : Any = UNetaDModel.from_pretrained("""diffusers/consistency_models""" , subfolder="""diffusers_cd_imagenet64_l2""" )
_A : Optional[int] = CMStochasticIterativeScheduler(
num_train_timesteps=40 , sigma_min=0.002 , sigma_max=80.0 , )
_A : Dict = ConsistencyModelPipeline(unet=_a , scheduler=_a )
pipe.to(torch_device=_a , torch_dtype=torch.floataa )
pipe.set_progress_bar_config(disable=_a )
_A : int = self.get_inputs(get_fixed_latents=_a , device=_a )
# Ensure usage of flash attention in torch 2.0
with sdp_kernel(enable_flash=_a , enable_math=_a , enable_mem_efficient=_a ):
_A : Union[str, Any] = pipe(**_a ).images
assert image.shape == (1, 64, 64, 3)
_A : Optional[Any] = image[0, -3:, -3:, -1]
_A : Dict = np.array([0.1875, 0.1428, 0.1289, 0.2151, 0.2092, 0.1477, 0.1877, 0.1641, 0.1353] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-3
@require_torch_a
def a__ ( self ) -> Optional[int]:
_A : int = UNetaDModel.from_pretrained("""diffusers/consistency_models""" , subfolder="""diffusers_cd_imagenet64_l2""" )
_A : Any = CMStochasticIterativeScheduler(
num_train_timesteps=40 , sigma_min=0.002 , sigma_max=80.0 , )
_A : str = ConsistencyModelPipeline(unet=_a , scheduler=_a )
pipe.to(torch_device=_a , torch_dtype=torch.floataa )
pipe.set_progress_bar_config(disable=_a )
_A : Any = self.get_inputs(get_fixed_latents=_a , device=_a )
_A : Dict = 1
_A : Union[str, Any] = None
# Ensure usage of flash attention in torch 2.0
with sdp_kernel(enable_flash=_a , enable_math=_a , enable_mem_efficient=_a ):
_A : int = pipe(**_a ).images
assert image.shape == (1, 64, 64, 3)
_A : Optional[int] = image[0, -3:, -3:, -1]
_A : Dict = np.array([0.1663, 0.1948, 0.2275, 0.1680, 0.1204, 0.1245, 0.1858, 0.1338, 0.2095] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-3
| 26 |
from __future__ import annotations
import unittest
from transformers import is_tf_available
from transformers.testing_utils import require_sentencepiece, require_tf, require_tokenizers, slow
if is_tf_available():
import numpy as np
import tensorflow as tf
from transformers import TFCamembertModel
@require_tf
@require_sentencepiece
@require_tokenizers
class lowercase ( unittest.TestCase ):
@slow
def a__ ( self ) -> Any:
_A : Tuple = TFCamembertModel.from_pretrained("""jplu/tf-camembert-base""" )
_A : List[Any] = tf.convert_to_tensor(
[[5, 121, 11, 660, 16, 730, 2_5543, 110, 83, 6]] , dtype=tf.intaa , ) # J'aime le camembert !"
_A : List[str] = model(_a )["""last_hidden_state"""]
_A : Union[str, Any] = tf.TensorShape((1, 10, 768) )
self.assertEqual(output.shape , _a )
# compare the actual values for a slice.
_A : List[Any] = tf.convert_to_tensor(
[[[-0.0254, 0.0235, 0.1027], [0.0606, -0.1811, -0.0418], [-0.1561, -0.1127, 0.2687]]] , dtype=tf.floataa , )
# camembert = torch.hub.load('pytorch/fairseq', 'camembert.v0')
# camembert.eval()
# expected_slice = roberta.model.forward(input_ids)[0][:, :3, :3].detach()
self.assertTrue(np.allclose(output[:, :3, :3].numpy() , expected_slice.numpy() , atol=1e-4 ) )
| 26 | 1 |
import io
import itertools
import json
from dataclasses import dataclass
from typing import Optional
import pyarrow as pa
import pyarrow.json as paj
import datasets
from datasets.table import table_cast
from datasets.utils.file_utils import readline
_snake_case = datasets.utils.logging.get_logger(__name__)
@dataclass
class lowercase ( datasets.BuilderConfig ):
_a = None
_a = "utf-8"
_a = None
_a = None
_a = True # deprecated
_a = None # deprecated
_a = 1_0 << 2_0 # 10MB
_a = None
class lowercase ( datasets.ArrowBasedBuilder ):
_a = JsonConfig
def a__ ( self ) -> Union[str, Any]:
if self.config.block_size is not None:
logger.warning("""The JSON loader parameter `block_size` is deprecated. Please use `chunksize` instead""" )
_A : Optional[Any] = self.config.block_size
if self.config.use_threads is not True:
logger.warning(
"""The JSON loader parameter `use_threads` is deprecated and doesn't have any effect anymore.""" )
if self.config.newlines_in_values is not None:
raise ValueError("""The JSON loader parameter `newlines_in_values` is no longer supported""" )
return datasets.DatasetInfo(features=self.config.features )
def a__ ( self , _a ) -> Union[str, Any]:
if not self.config.data_files:
raise ValueError(F'''At least one data file must be specified, but got data_files={self.config.data_files}''' )
_A : Optional[Any] = dl_manager.download_and_extract(self.config.data_files )
if isinstance(_a , (str, list, tuple) ):
_A : List[str] = data_files
if isinstance(_a , _a ):
_A : Optional[int] = [files]
_A : str = [dl_manager.iter_files(_a ) for file in files]
return [datasets.SplitGenerator(name=datasets.Split.TRAIN , gen_kwargs={"""files""": files} )]
_A : Tuple = []
for split_name, files in data_files.items():
if isinstance(_a , _a ):
_A : Optional[int] = [files]
_A : Tuple = [dl_manager.iter_files(_a ) for file in files]
splits.append(datasets.SplitGenerator(name=_a , gen_kwargs={"""files""": files} ) )
return splits
def a__ ( self , _a ) -> pa.Table:
if self.config.features is not None:
# adding missing columns
for column_name in set(self.config.features ) - set(pa_table.column_names ):
_A : List[Any] = self.config.features.arrow_schema.field(_a ).type
_A : str = pa_table.append_column(_a , pa.array([None] * len(_a ) , type=_a ) )
# more expensive cast to support nested structures with keys in a different order
# allows str <-> int/float or str to Audio for example
_A : Optional[Any] = table_cast(_a , self.config.features.arrow_schema )
return pa_table
def a__ ( self , _a ) -> Optional[int]:
for file_idx, file in enumerate(itertools.chain.from_iterable(_a ) ):
# If the file is one json object and if we need to look at the list of items in one specific field
if self.config.field is not None:
with open(_a , encoding=self.config.encoding , errors=self.config.encoding_errors ) as f:
_A : Any = json.load(_a )
# We keep only the field we are interested in
_A : Optional[int] = dataset[self.config.field]
# We accept two format: a list of dicts or a dict of lists
if isinstance(_a , (list, tuple) ):
_A : Union[str, Any] = set().union(*[row.keys() for row in dataset] )
_A : Union[str, Any] = {col: [row.get(_a ) for row in dataset] for col in keys}
else:
_A : List[str] = dataset
_A : Optional[Any] = pa.Table.from_pydict(_a )
yield file_idx, self._cast_table(_a )
# If the file has one json object per line
else:
with open(_a , """rb""" ) as f:
_A : Dict = 0
# Use block_size equal to the chunk size divided by 32 to leverage multithreading
# Set a default minimum value of 16kB if the chunk size is really small
_A : List[str] = max(self.config.chunksize // 32 , 16 << 10 )
_A : Any = (
self.config.encoding_errors if self.config.encoding_errors is not None else """strict"""
)
while True:
_A : List[Any] = f.read(self.config.chunksize )
if not batch:
break
# Finish current line
try:
batch += f.readline()
except (AttributeError, io.UnsupportedOperation):
batch += readline(_a )
# PyArrow only accepts utf-8 encoded bytes
if self.config.encoding != "utf-8":
_A : List[str] = batch.decode(self.config.encoding , errors=_a ).encode("""utf-8""" )
try:
while True:
try:
_A : Union[str, Any] = paj.read_json(
io.BytesIO(_a ) , read_options=paj.ReadOptions(block_size=_a ) )
break
except (pa.ArrowInvalid, pa.ArrowNotImplementedError) as e:
if (
isinstance(_a , pa.ArrowInvalid )
and "straddling" not in str(_a )
or block_size > len(_a )
):
raise
else:
# Increase the block size in case it was too small.
# The block size will be reset for the next file.
logger.debug(
F'''Batch of {len(_a )} bytes couldn\'t be parsed with block_size={block_size}. Retrying with block_size={block_size * 2}.''' )
block_size *= 2
except pa.ArrowInvalid as e:
try:
with open(
_a , encoding=self.config.encoding , errors=self.config.encoding_errors ) as f:
_A : List[str] = json.load(_a )
except json.JSONDecodeError:
logger.error(F'''Failed to read file \'{file}\' with error {type(_a )}: {e}''' )
raise e
# If possible, parse the file as a list of json objects and exit the loop
if isinstance(_a , _a ): # list is the only sequence type supported in JSON
try:
_A : str = set().union(*[row.keys() for row in dataset] )
_A : Dict = {col: [row.get(_a ) for row in dataset] for col in keys}
_A : List[Any] = pa.Table.from_pydict(_a )
except (pa.ArrowInvalid, AttributeError) as e:
logger.error(F'''Failed to read file \'{file}\' with error {type(_a )}: {e}''' )
raise ValueError(F'''Not able to read records in the JSON file at {file}.''' ) from None
yield file_idx, self._cast_table(_a )
break
else:
logger.error(F'''Failed to read file \'{file}\' with error {type(_a )}: {e}''' )
raise ValueError(
F'''Not able to read records in the JSON file at {file}. '''
F'''You should probably indicate the field of the JSON file containing your records. '''
F'''This JSON file contain the following fields: {str(list(dataset.keys() ) )}. '''
F'''Select the correct one and provide it as `field=\'XXX\'` to the dataset loading method. ''' ) from None
# Uncomment for debugging (will print the Arrow table size and elements)
# logger.warning(f"pa_table: {pa_table} num rows: {pa_table.num_rows}")
# logger.warning('\n'.join(str(pa_table.slice(i, 1).to_pydict()) for i in range(pa_table.num_rows)))
yield (file_idx, batch_idx), self._cast_table(_a )
batch_idx += 1
| 26 |
from dataclasses import dataclass
from typing import List, Optional, Union
import numpy as np
import PIL
import torch
from transformers import CLIPImageProcessor, CLIPVisionModel
from ...models import PriorTransformer
from ...pipelines import DiffusionPipeline
from ...schedulers import HeunDiscreteScheduler
from ...utils import (
BaseOutput,
is_accelerate_available,
logging,
randn_tensor,
replace_example_docstring,
)
from .renderer import ShapERenderer
_snake_case = logging.get_logger(__name__) # pylint: disable=invalid-name
_snake_case = "\n Examples:\n ```py\n >>> from PIL import Image\n >>> import torch\n >>> from diffusers import DiffusionPipeline\n >>> from diffusers.utils import export_to_gif, load_image\n\n >>> device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n\n >>> repo = \"openai/shap-e-img2img\"\n >>> pipe = DiffusionPipeline.from_pretrained(repo, torch_dtype=torch.float16)\n >>> pipe = pipe.to(device)\n\n >>> guidance_scale = 3.0\n >>> image_url = \"https://hf.co/datasets/diffusers/docs-images/resolve/main/shap-e/corgi.png\"\n >>> image = load_image(image_url).convert(\"RGB\")\n\n >>> images = pipe(\n ... image,\n ... guidance_scale=guidance_scale,\n ... num_inference_steps=64,\n ... frame_size=256,\n ... ).images\n\n >>> gif_path = export_to_gif(images[0], \"corgi_3d.gif\")\n ```\n"
@dataclass
class lowercase ( UpperCamelCase__ ):
_a = 42
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a , _a , _a , ) -> List[Any]:
super().__init__()
self.register_modules(
prior=_a , image_encoder=_a , image_processor=_a , scheduler=_a , renderer=_a , )
def a__ ( self , _a , _a , _a , _a , _a , _a ) -> str:
if latents is None:
_A : str = randn_tensor(_a , generator=_a , device=_a , dtype=_a )
else:
if latents.shape != shape:
raise ValueError(F'''Unexpected latents shape, got {latents.shape}, expected {shape}''' )
_A : Union[str, Any] = latents.to(_a )
_A : int = latents * scheduler.init_noise_sigma
return latents
def a__ ( self , _a=0 ) -> Optional[Any]:
if is_accelerate_available():
from accelerate import cpu_offload
else:
raise ImportError("""Please install accelerate via `pip install accelerate`""" )
_A : str = torch.device(F'''cuda:{gpu_id}''' )
_A : Any = [self.image_encoder, self.prior]
for cpu_offloaded_model in models:
if cpu_offloaded_model is not None:
cpu_offload(_a , _a )
@property
def a__ ( self ) -> List[Any]:
if self.device != torch.device("""meta""" ) or not hasattr(self.image_encoder , """_hf_hook""" ):
return self.device
for module in self.image_encoder.modules():
if (
hasattr(_a , """_hf_hook""" )
and hasattr(module._hf_hook , """execution_device""" )
and module._hf_hook.execution_device is not None
):
return torch.device(module._hf_hook.execution_device )
return self.device
def a__ ( self , _a , _a , _a , _a , ) -> Tuple:
if isinstance(_a , _a ) and isinstance(image[0] , torch.Tensor ):
_A : int = torch.cat(_a , axis=0 ) if image[0].ndim == 4 else torch.stack(_a , axis=0 )
if not isinstance(_a , torch.Tensor ):
_A : Dict = self.image_processor(_a , return_tensors="""pt""" ).pixel_values[0].unsqueeze(0 )
_A : int = image.to(dtype=self.image_encoder.dtype , device=_a )
_A : List[Any] = self.image_encoder(_a )["""last_hidden_state"""]
_A : List[Any] = image_embeds[:, 1:, :].contiguous() # batch_size, dim, 256
_A : Dict = image_embeds.repeat_interleave(_a , dim=0 )
if do_classifier_free_guidance:
_A : str = torch.zeros_like(_a )
# For classifier free guidance, we need to do two forward passes.
# Here we concatenate the unconditional and text embeddings into a single batch
# to avoid doing two forward passes
_A : List[str] = torch.cat([negative_image_embeds, image_embeds] )
return image_embeds
@torch.no_grad()
@replace_example_docstring(_a )
def __call__( self , _a , _a = 1 , _a = 25 , _a = None , _a = None , _a = 4.0 , _a = 64 , _a = "pil" , _a = True , ) -> Union[str, Any]:
if isinstance(_a , PIL.Image.Image ):
_A : List[Any] = 1
elif isinstance(_a , torch.Tensor ):
_A : Any = image.shape[0]
elif isinstance(_a , _a ) and isinstance(image[0] , (torch.Tensor, PIL.Image.Image) ):
_A : Union[str, Any] = len(_a )
else:
raise ValueError(
F'''`image` has to be of type `PIL.Image.Image`, `torch.Tensor`, `List[PIL.Image.Image]` or `List[torch.Tensor]` but is {type(_a )}''' )
_A : Optional[int] = self._execution_device
_A : Tuple = batch_size * num_images_per_prompt
_A : List[Any] = guidance_scale > 1.0
_A : Optional[Any] = self._encode_image(_a , _a , _a , _a )
# prior
self.scheduler.set_timesteps(_a , device=_a )
_A : Optional[int] = self.scheduler.timesteps
_A : List[str] = self.prior.config.num_embeddings
_A : int = self.prior.config.embedding_dim
_A : Optional[Any] = self.prepare_latents(
(batch_size, num_embeddings * embedding_dim) , image_embeds.dtype , _a , _a , _a , self.scheduler , )
# YiYi notes: for testing only to match ldm, we can directly create a latents with desired shape: batch_size, num_embeddings, embedding_dim
_A : List[Any] = latents.reshape(latents.shape[0] , _a , _a )
for i, t in enumerate(self.progress_bar(_a ) ):
# expand the latents if we are doing classifier free guidance
_A : List[str] = torch.cat([latents] * 2 ) if do_classifier_free_guidance else latents
_A : int = self.scheduler.scale_model_input(_a , _a )
_A : Tuple = self.prior(
_a , timestep=_a , proj_embedding=_a , ).predicted_image_embedding
# remove the variance
_A , _A : Optional[Any] = noise_pred.split(
scaled_model_input.shape[2] , dim=2 ) # batch_size, num_embeddings, embedding_dim
if do_classifier_free_guidance is not None:
_A , _A : Dict = noise_pred.chunk(2 )
_A : Tuple = noise_pred_uncond + guidance_scale * (noise_pred - noise_pred_uncond)
_A : int = self.scheduler.step(
_a , timestep=_a , sample=_a , ).prev_sample
if output_type == "latent":
return ShapEPipelineOutput(images=_a )
_A : List[str] = []
for i, latent in enumerate(_a ):
print()
_A : List[str] = self.renderer.decode(
latent[None, :] , _a , size=_a , ray_batch_size=4096 , n_coarse_samples=64 , n_fine_samples=128 , )
images.append(_a )
_A : List[Any] = torch.stack(_a )
if output_type not in ["np", "pil"]:
raise ValueError(F'''Only the output types `pil` and `np` are supported not output_type={output_type}''' )
_A : List[str] = images.cpu().numpy()
if output_type == "pil":
_A : List[Any] = [self.numpy_to_pil(_a ) for image in images]
# Offload last model to CPU
if hasattr(self , """final_offload_hook""" ) and self.final_offload_hook is not None:
self.final_offload_hook.offload()
if not return_dict:
return (images,)
return ShapEPipelineOutput(images=_a )
| 26 | 1 |
from __future__ import annotations
import copy
import inspect
import unittest
import numpy as np
from transformers import is_tf_available, is_vision_available
from transformers.models.auto import get_values
from transformers.testing_utils import require_tf, slow
from transformers.utils import cached_property
from ...test_configuration_common import ConfigTester
from ...test_modeling_tf_common import TFModelTesterMixin, floats_tensor, ids_tensor, random_attention_mask
from ...test_pipeline_mixin import PipelineTesterMixin
if is_tf_available():
import tensorflow as tf
from transformers import (
TF_LAYOUTLMV3_PRETRAINED_MODEL_ARCHIVE_LIST,
TF_MODEL_FOR_MULTIPLE_CHOICE_MAPPING,
TF_MODEL_FOR_QUESTION_ANSWERING_MAPPING,
TF_MODEL_FOR_SEQUENCE_CLASSIFICATION_MAPPING,
TF_MODEL_FOR_TOKEN_CLASSIFICATION_MAPPING,
LayoutLMvaConfig,
TFLayoutLMvaForQuestionAnswering,
TFLayoutLMvaForSequenceClassification,
TFLayoutLMvaForTokenClassification,
TFLayoutLMvaModel,
)
if is_vision_available():
from PIL import Image
from transformers import LayoutLMvaImageProcessor
class lowercase :
def __init__( self , _a , _a=2 , _a=3 , _a=4 , _a=2 , _a=7 , _a=True , _a=True , _a=True , _a=True , _a=99 , _a=36 , _a=2 , _a=4 , _a=37 , _a="gelu" , _a=0.1 , _a=0.1 , _a=512 , _a=16 , _a=2 , _a=0.02 , _a=6 , _a=6 , _a=3 , _a=4 , _a=None , _a=1000 , ) -> int:
_A : Tuple = parent
_A : int = batch_size
_A : Union[str, Any] = num_channels
_A : int = image_size
_A : Tuple = patch_size
_A : Union[str, Any] = is_training
_A : List[Any] = use_input_mask
_A : Tuple = use_token_type_ids
_A : Optional[int] = use_labels
_A : Any = vocab_size
_A : Union[str, Any] = hidden_size
_A : Dict = num_hidden_layers
_A : Dict = num_attention_heads
_A : List[Any] = intermediate_size
_A : Optional[Any] = hidden_act
_A : Optional[Any] = hidden_dropout_prob
_A : Dict = attention_probs_dropout_prob
_A : Any = max_position_embeddings
_A : int = type_vocab_size
_A : Tuple = type_sequence_label_size
_A : Optional[int] = initializer_range
_A : List[Any] = coordinate_size
_A : Dict = shape_size
_A : Tuple = num_labels
_A : str = num_choices
_A : str = scope
_A : Optional[Any] = range_bbox
# LayoutLMv3's sequence length equals the number of text tokens + number of patches + 1 (we add 1 for the CLS token)
_A : Union[str, Any] = text_seq_length
_A : Optional[Any] = (image_size // patch_size) ** 2 + 1
_A : List[str] = self.text_seq_length + self.image_seq_length
def a__ ( self ) -> str:
_A : Union[str, Any] = ids_tensor([self.batch_size, self.text_seq_length] , self.vocab_size )
_A : Any = ids_tensor([self.batch_size, self.text_seq_length, 4] , self.range_bbox )
_A : Dict = bbox.numpy()
# Ensure that bbox is legal
for i in range(bbox.shape[0] ):
for j in range(bbox.shape[1] ):
if bbox[i, j, 3] < bbox[i, j, 1]:
_A : Union[str, Any] = bbox[i, j, 3]
_A : Optional[int] = bbox[i, j, 1]
_A : Dict = tmp_coordinate
if bbox[i, j, 2] < bbox[i, j, 0]:
_A : Union[str, Any] = bbox[i, j, 2]
_A : Dict = bbox[i, j, 0]
_A : Union[str, Any] = tmp_coordinate
_A : Any = tf.constant(_a )
_A : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
_A : List[Any] = None
if self.use_input_mask:
_A : List[str] = random_attention_mask([self.batch_size, self.text_seq_length] )
_A : List[Any] = None
if self.use_token_type_ids:
_A : str = ids_tensor([self.batch_size, self.text_seq_length] , self.type_vocab_size )
_A : Union[str, Any] = None
_A : Union[str, Any] = None
if self.use_labels:
_A : int = ids_tensor([self.batch_size] , self.type_sequence_label_size )
_A : List[Any] = ids_tensor([self.batch_size, self.text_seq_length] , self.num_labels )
_A : Optional[int] = LayoutLMvaConfig(
vocab_size=self.vocab_size , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , initializer_range=self.initializer_range , coordinate_size=self.coordinate_size , shape_size=self.shape_size , input_size=self.image_size , patch_size=self.patch_size , )
return config, input_ids, bbox, pixel_values, token_type_ids, input_mask, sequence_labels, token_labels
def a__ ( self , _a , _a , _a , _a , _a , _a ) -> Optional[int]:
_A : str = TFLayoutLMvaModel(config=_a )
# text + image
_A : List[str] = model(_a , pixel_values=_a , training=_a )
_A : Union[str, Any] = model(
_a , bbox=_a , pixel_values=_a , attention_mask=_a , token_type_ids=_a , training=_a , )
_A : int = model(_a , bbox=_a , pixel_values=_a , training=_a )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) )
# text only
_A : Dict = model(_a , training=_a )
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.text_seq_length, self.hidden_size) )
# image only
_A : Union[str, Any] = model({"""pixel_values""": pixel_values} , training=_a )
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.image_seq_length, self.hidden_size) )
def a__ ( self , _a , _a , _a , _a , _a , _a , _a ) -> int:
_A : str = self.num_labels
_A : List[str] = TFLayoutLMvaForSequenceClassification(config=_a )
_A : int = model(
_a , bbox=_a , pixel_values=_a , attention_mask=_a , token_type_ids=_a , labels=_a , training=_a , )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) )
def a__ ( self , _a , _a , _a , _a , _a , _a , _a ) -> Any:
_A : str = self.num_labels
_A : Any = TFLayoutLMvaForTokenClassification(config=_a )
_A : Union[str, Any] = model(
_a , bbox=_a , pixel_values=_a , attention_mask=_a , token_type_ids=_a , labels=_a , training=_a , )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.text_seq_length, self.num_labels) )
def a__ ( self , _a , _a , _a , _a , _a , _a , _a ) -> Optional[int]:
_A : int = 2
_A : Dict = TFLayoutLMvaForQuestionAnswering(config=_a )
_A : Tuple = model(
_a , bbox=_a , pixel_values=_a , attention_mask=_a , token_type_ids=_a , start_positions=_a , end_positions=_a , training=_a , )
self.parent.assertEqual(result.start_logits.shape , (self.batch_size, self.seq_length) )
self.parent.assertEqual(result.end_logits.shape , (self.batch_size, self.seq_length) )
def a__ ( self ) -> Union[str, Any]:
_A : int = self.prepare_config_and_inputs()
((_A) , (_A) , (_A) , (_A) , (_A) , (_A) , (_A) , (_A)) : Any = config_and_inputs
_A : List[Any] = {
"""input_ids""": input_ids,
"""bbox""": bbox,
"""pixel_values""": pixel_values,
"""token_type_ids""": token_type_ids,
"""attention_mask""": input_mask,
}
return config, inputs_dict
@require_tf
class lowercase ( UpperCamelCase__,UpperCamelCase__,unittest.TestCase ):
_a = (
(
TFLayoutLMvaModel,
TFLayoutLMvaForQuestionAnswering,
TFLayoutLMvaForSequenceClassification,
TFLayoutLMvaForTokenClassification,
)
if is_tf_available()
else ()
)
_a = (
{"document-question-answering": TFLayoutLMvaForQuestionAnswering, "feature-extraction": TFLayoutLMvaModel}
if is_tf_available()
else {}
)
_a = False
_a = False
_a = False
def a__ ( self , _a , _a , _a , _a , _a ) -> Optional[int]:
return True
def a__ ( self , _a , _a , _a=False ) -> dict:
_A : int = copy.deepcopy(_a )
if model_class in get_values(_a ):
_A : List[Any] = {
k: tf.tile(tf.expand_dims(_a , 1 ) , (1, self.model_tester.num_choices) + (1,) * (v.ndim - 1) )
if isinstance(_a , tf.Tensor ) and v.ndim > 0
else v
for k, v in inputs_dict.items()
}
if return_labels:
if model_class in get_values(_a ):
_A : List[str] = tf.ones(self.model_tester.batch_size , dtype=tf.intaa )
elif model_class in get_values(_a ):
_A : Dict = tf.zeros(self.model_tester.batch_size , dtype=tf.intaa )
_A : List[Any] = tf.zeros(self.model_tester.batch_size , dtype=tf.intaa )
elif model_class in get_values(_a ):
_A : str = tf.zeros(self.model_tester.batch_size , dtype=tf.intaa )
elif model_class in get_values(_a ):
_A : Tuple = tf.zeros(
(self.model_tester.batch_size, self.model_tester.text_seq_length) , dtype=tf.intaa )
return inputs_dict
def a__ ( self ) -> List[str]:
_A : List[str] = TFLayoutLMvaModelTester(self )
_A : Optional[Any] = ConfigTester(self , config_class=_a , hidden_size=37 )
def a__ ( self ) -> Optional[Any]:
self.config_tester.run_common_tests()
def a__ ( self ) -> List[str]:
_A , _A : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
_A : Dict = model_class(_a )
if getattr(_a , """hf_compute_loss""" , _a ):
# The number of elements in the loss should be the same as the number of elements in the label
_A : int = self._prepare_for_class(inputs_dict.copy() , _a , return_labels=_a )
_A : int = prepared_for_class[
sorted(prepared_for_class.keys() - inputs_dict.keys() , reverse=_a )[0]
]
_A : int = added_label.shape.as_list()[:1]
# Test that model correctly compute the loss with kwargs
_A : List[str] = self._prepare_for_class(inputs_dict.copy() , _a , return_labels=_a )
_A : Union[str, Any] = prepared_for_class.pop("""input_ids""" )
_A : Tuple = model(_a , **_a )[0]
self.assertTrue(loss.shape.as_list() == expected_loss_size or loss.shape.as_list() == [1] )
# Test that model correctly compute the loss when we mask some positions
_A : Tuple = self._prepare_for_class(inputs_dict.copy() , _a , return_labels=_a )
_A : Any = prepared_for_class.pop("""input_ids""" )
if "labels" in prepared_for_class:
_A : int = prepared_for_class["""labels"""].numpy()
if len(labels.shape ) > 1 and labels.shape[1] != 1:
_A : Tuple = -100
_A : Dict = tf.convert_to_tensor(_a )
_A : Any = model(_a , **_a )[0]
self.assertTrue(loss.shape.as_list() == expected_loss_size or loss.shape.as_list() == [1] )
self.assertTrue(not np.any(np.isnan(loss.numpy() ) ) )
# Test that model correctly compute the loss with a dict
_A : List[str] = self._prepare_for_class(inputs_dict.copy() , _a , return_labels=_a )
_A : int = model(_a )[0]
self.assertTrue(loss.shape.as_list() == expected_loss_size or loss.shape.as_list() == [1] )
# Test that model correctly compute the loss with a tuple
_A : List[str] = self._prepare_for_class(inputs_dict.copy() , _a , return_labels=_a )
# Get keys that were added with the _prepare_for_class function
_A : str = prepared_for_class.keys() - inputs_dict.keys()
_A : Dict = inspect.signature(model.call ).parameters
_A : Dict = list(signature.keys() )
# Create a dictionary holding the location of the tensors in the tuple
_A : Dict = {0: """input_ids"""}
for label_key in label_keys:
_A : str = signature_names.index(_a )
_A : int = label_key
_A : Dict = sorted(tuple_index_mapping.items() )
# Initialize a list with their default values, update the values and convert to a tuple
_A : List[str] = []
for name in signature_names:
if name != "kwargs":
list_input.append(signature[name].default )
for index, value in sorted_tuple_index_mapping:
_A : Tuple = prepared_for_class[value]
_A : int = tuple(_a )
# Send to model
_A : List[str] = model(tuple_input[:-1] )[0]
self.assertTrue(loss.shape.as_list() == expected_loss_size or loss.shape.as_list() == [1] )
def a__ ( self ) -> Optional[int]:
(
(
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) ,
) : Optional[int] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(_a , _a , _a , _a , _a , _a )
def a__ ( self ) -> Any:
(
(
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) ,
) : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
for type in ["absolute", "relative_key", "relative_key_query"]:
_A : Optional[Any] = type
self.model_tester.create_and_check_model(_a , _a , _a , _a , _a , _a )
def a__ ( self ) -> Tuple:
(
(
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) ,
) : List[Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_sequence_classification(
_a , _a , _a , _a , _a , _a , _a )
def a__ ( self ) -> Tuple:
(
(
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) ,
) : List[Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_token_classification(
_a , _a , _a , _a , _a , _a , _a )
def a__ ( self ) -> List[str]:
(
(
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) , (
_A
) ,
) : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_question_answering(
_a , _a , _a , _a , _a , _a , _a )
@slow
def a__ ( self ) -> str:
for model_name in TF_LAYOUTLMV3_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
_A : Tuple = TFLayoutLMvaModel.from_pretrained(_a )
self.assertIsNotNone(_a )
def lowerCAmelCase_ ( ):
_A : List[Any] = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_tf
class lowercase ( unittest.TestCase ):
@cached_property
def a__ ( self ) -> Tuple:
return LayoutLMvaImageProcessor(apply_ocr=_a ) if is_vision_available() else None
@slow
def a__ ( self ) -> Union[str, Any]:
_A : Optional[int] = TFLayoutLMvaModel.from_pretrained("""microsoft/layoutlmv3-base""" )
_A : Tuple = self.default_image_processor
_A : int = prepare_img()
_A : List[Any] = image_processor(images=_a , return_tensors="""tf""" ).pixel_values
_A : str = tf.constant([[1, 2]] )
_A : int = tf.expand_dims(tf.constant([[1, 2, 3, 4], [5, 6, 7, 8]] ) , axis=0 )
# forward pass
_A : Tuple = model(input_ids=_a , bbox=_a , pixel_values=_a , training=_a )
# verify the logits
_A : int = (1, 199, 768)
self.assertEqual(outputs.last_hidden_state.shape , _a )
_A : List[Any] = tf.constant(
[[-0.0529, 0.3618, 0.1632], [-0.1587, -0.1667, -0.0400], [-0.1557, -0.1671, -0.0505]] )
self.assertTrue(np.allclose(outputs.last_hidden_state[0, :3, :3] , _a , atol=1e-4 ) )
| 26 |
import argparse
import collections
import json
from pathlib import Path
import requests
import torch
import yaml
from huggingface_hub import hf_hub_download
from PIL import Image
from transformers import (
MobileViTImageProcessor,
MobileViTVaConfig,
MobileViTVaForImageClassification,
MobileViTVaForSemanticSegmentation,
)
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
def lowerCAmelCase_ ( snake_case_ ):
print("""Loading config file...""" )
def flatten_yaml_as_dict(snake_case_,snake_case_="",snake_case_="." ):
_A : Union[str, Any] = []
for k, v in d.items():
_A : Optional[int] = parent_key + sep + k if parent_key else k
if isinstance(snake_case_,collections.abc.MutableMapping ):
items.extend(flatten_yaml_as_dict(snake_case_,snake_case_,sep=snake_case_ ).items() )
else:
items.append((new_key, v) )
return dict(snake_case_ )
_A : List[Any] = argparse.Namespace()
with open(snake_case_,"""r""" ) as yaml_file:
try:
_A : List[Any] = yaml.load(snake_case_,Loader=yaml.FullLoader )
_A : Optional[int] = flatten_yaml_as_dict(snake_case_ )
for k, v in flat_cfg.items():
setattr(snake_case_,snake_case_,snake_case_ )
except yaml.YAMLError as exc:
logger.error("""Error while loading config file: {}. Error message: {}""".format(snake_case_,str(snake_case_ ) ) )
return config
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Optional[Any] = MobileViTVaConfig()
_A : Tuple = False
# dataset
if task_name.startswith("""imagenet1k_""" ):
_A : Dict = 1000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : int = 384
else:
_A : int = 256
_A : List[str] = """imagenet-1k-id2label.json"""
elif task_name.startswith("""imagenet21k_to_1k_""" ):
_A : Union[str, Any] = 21000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : str = 384
else:
_A : List[Any] = 256
_A : List[str] = """imagenet-22k-id2label.json"""
elif task_name.startswith("""ade20k_""" ):
_A : int = 151
_A : int = 512
_A : Optional[int] = """ade20k-id2label.json"""
_A : Any = True
elif task_name.startswith("""voc_""" ):
_A : List[Any] = 21
_A : Dict = 512
_A : Dict = """pascal-voc-id2label.json"""
_A : int = True
# orig_config
_A : Any = load_orig_config_file(snake_case_ )
assert getattr(snake_case_,"""model.classification.name""",-1 ) == "mobilevit_v2", "Invalid model"
_A : List[Any] = getattr(snake_case_,"""model.classification.mitv2.width_multiplier""",1.0 )
assert (
getattr(snake_case_,"""model.classification.mitv2.attn_norm_layer""",-1 ) == "layer_norm_2d"
), "Norm layers other than layer_norm_2d is not supported"
_A : str = getattr(snake_case_,"""model.classification.activation.name""","""swish""" )
# config.image_size == getattr(orig_config, 'sampler.bs.crop_size_width', 256)
if is_segmentation_model:
_A : Optional[int] = getattr(snake_case_,"""model.segmentation.output_stride""",16 )
if "_deeplabv3" in task_name:
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_rates""",[12, 24, 36] )
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_out_channels""",512 )
_A : str = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_dropout""",0.1 )
# id2label
_A : List[Any] = """huggingface/label-files"""
_A : List[Any] = json.load(open(hf_hub_download(snake_case_,snake_case_,repo_type="""dataset""" ),"""r""" ) )
_A : str = {int(snake_case_ ): v for k, v in idalabel.items()}
_A : str = idalabel
_A : Dict = {v: k for k, v in idalabel.items()}
return config
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Any = dct.pop(snake_case_ )
_A : Union[str, Any] = val
def lowerCAmelCase_ ( snake_case_,snake_case_=False ):
if base_model:
_A : Optional[int] = """"""
else:
_A : Dict = """mobilevitv2."""
_A : int = []
for k in state_dict.keys():
if k[:8] == "encoder.":
_A : Any = k[8:]
else:
_A : List[str] = k
if ".block." in k:
_A : Any = k_new.replace(""".block.""",""".""" )
if ".conv." in k:
_A : List[Any] = k_new.replace(""".conv.""",""".convolution.""" )
if ".norm." in k:
_A : Any = k_new.replace(""".norm.""",""".normalization.""" )
if "conv_1." in k:
_A : int = k_new.replace("""conv_1.""",f'''{model_prefix}conv_stem.''' )
for i in [1, 2]:
if f'''layer_{i}.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.''',f'''{model_prefix}encoder.layer.{i-1}.layer.''' )
if ".exp_1x1." in k:
_A : Tuple = k_new.replace(""".exp_1x1.""",""".expand_1x1.""" )
if ".red_1x1." in k:
_A : Optional[int] = k_new.replace(""".red_1x1.""",""".reduce_1x1.""" )
for i in [3, 4, 5]:
if f'''layer_{i}.0.''' in k:
_A : Optional[int] = k_new.replace(f'''layer_{i}.0.''',f'''{model_prefix}encoder.layer.{i-1}.downsampling_layer.''' )
if f'''layer_{i}.1.local_rep.0.''' in k:
_A : Union[str, Any] = k_new.replace(f'''layer_{i}.1.local_rep.0.''',f'''{model_prefix}encoder.layer.{i-1}.conv_kxk.''' )
if f'''layer_{i}.1.local_rep.1.''' in k:
_A : str = k_new.replace(f'''layer_{i}.1.local_rep.1.''',f'''{model_prefix}encoder.layer.{i-1}.conv_1x1.''' )
for i in [3, 4, 5]:
if i == 3:
_A : Optional[int] = [0, 1]
elif i == 4:
_A : Union[str, Any] = [0, 1, 2, 3]
elif i == 5:
_A : Optional[Any] = [0, 1, 2]
for j in j_in:
if f'''layer_{i}.1.global_rep.{j}.''' in k:
_A : Union[str, Any] = k_new.replace(
f'''layer_{i}.1.global_rep.{j}.''',f'''{model_prefix}encoder.layer.{i-1}.transformer.layer.{j}.''' )
if f'''layer_{i}.1.global_rep.{j+1}.''' in k:
_A : List[str] = k_new.replace(
f'''layer_{i}.1.global_rep.{j+1}.''',f'''{model_prefix}encoder.layer.{i-1}.layernorm.''' )
if f'''layer_{i}.1.conv_proj.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.1.conv_proj.''',f'''{model_prefix}encoder.layer.{i-1}.conv_projection.''' )
if "pre_norm_attn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_attn.0.""","""layernorm_before.""" )
if "pre_norm_attn.1." in k:
_A : str = k_new.replace("""pre_norm_attn.1.""","""attention.""" )
if "pre_norm_ffn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_ffn.0.""","""layernorm_after.""" )
if "pre_norm_ffn.1." in k:
_A : Dict = k_new.replace("""pre_norm_ffn.1.""","""ffn.conv1.""" )
if "pre_norm_ffn.3." in k:
_A : List[str] = k_new.replace("""pre_norm_ffn.3.""","""ffn.conv2.""" )
if "classifier.1." in k:
_A : List[str] = k_new.replace("""classifier.1.""","""classifier.""" )
if "seg_head." in k:
_A : List[Any] = k_new.replace("""seg_head.""","""segmentation_head.""" )
if ".aspp_layer." in k:
_A : List[Any] = k_new.replace(""".aspp_layer.""",""".""" )
if ".aspp_pool." in k:
_A : Optional[Any] = k_new.replace(""".aspp_pool.""",""".""" )
rename_keys.append((k, k_new) )
return rename_keys
def lowerCAmelCase_ ( snake_case_ ):
_A : Tuple = []
for k in state_dict.keys():
if k.startswith("""seg_head.aux_head.""" ):
keys_to_ignore.append(snake_case_ )
for k in keys_to_ignore:
state_dict.pop(snake_case_,snake_case_ )
def lowerCAmelCase_ ( ):
_A : Dict = """http://images.cocodataset.org/val2017/000000039769.jpg"""
# url = "https://cdn.britannica.com/86/141086-050-9D7C75EE/Gulfstream-G450-business-jet-passengers.jpg"
_A : List[Any] = Image.open(requests.get(snake_case_,stream=snake_case_ ).raw )
return im
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : List[Any] = get_mobilevitva_config(snake_case_,snake_case_ )
# load original state_dict
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
# load huggingface model
if task_name.startswith("""ade20k_""" ) or task_name.startswith("""voc_""" ):
_A : Optional[Any] = MobileViTVaForSemanticSegmentation(snake_case_ ).eval()
_A : str = False
else:
_A : int = MobileViTVaForImageClassification(snake_case_ ).eval()
_A : List[Any] = False
# remove and rename some keys of load the original model
_A : List[Any] = checkpoint
remove_unused_keys(snake_case_ )
_A : Optional[Any] = create_rename_keys(snake_case_,base_model=snake_case_ )
for rename_key_src, rename_key_dest in rename_keys:
rename_key(snake_case_,snake_case_,snake_case_ )
# load modified state_dict
model.load_state_dict(snake_case_ )
# Check outputs on an image, prepared by MobileViTImageProcessor
_A : str = MobileViTImageProcessor(crop_size=config.image_size,size=config.image_size + 32 )
_A : List[Any] = image_processor(images=prepare_img(),return_tensors="""pt""" )
_A : Optional[Any] = model(**snake_case_ )
# verify classification model
if task_name.startswith("""imagenet""" ):
_A : List[Any] = outputs.logits
_A : Optional[int] = logits.argmax(-1 ).item()
print("""Predicted class:""",model.config.idalabel[predicted_class_idx] )
if task_name.startswith("""imagenet1k_256""" ) and config.width_multiplier == 1.0:
# expected_logits for base variant
_A : int = torch.tensor([-1.63_36e00, -7.32_04e-02, -5.18_83e-01] )
assert torch.allclose(logits[0, :3],snake_case_,atol=1e-4 )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ )
print(f'''Saving model {task_name} to {pytorch_dump_folder_path}''' )
model.save_pretrained(snake_case_ )
print(f'''Saving image processor to {pytorch_dump_folder_path}''' )
image_processor.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
"--task",
default="imagenet1k_256",
type=str,
help=(
"Name of the task for which the MobileViTV2 model you'd like to convert is trained on . "
"\n Classification (ImageNet-1k)\n - MobileViTV2 (256x256) : imagenet1k_256\n - MobileViTV2 (Trained on 256x256 and Finetuned on 384x384) : imagenet1k_384\n - MobileViTV2 (Trained on ImageNet-21k and Finetuned on ImageNet-1k 256x256) :\n imagenet21k_to_1k_256\n - MobileViTV2 (Trained on ImageNet-21k, Finetuned on ImageNet-1k 256x256, and Finetuned on\n ImageNet-1k 384x384) : imagenet21k_to_1k_384\n Segmentation\n - ADE20K Dataset : ade20k_deeplabv3\n - Pascal VOC 2012 Dataset: voc_deeplabv3\n "
),
choices=[
"imagenet1k_256",
"imagenet1k_384",
"imagenet21k_to_1k_256",
"imagenet21k_to_1k_384",
"ade20k_deeplabv3",
"voc_deeplabv3",
],
)
parser.add_argument(
"--orig_checkpoint_path", required=True, type=str, help="Path to the original state dict (.pt file)."
)
parser.add_argument("--orig_config_path", required=True, type=str, help="Path to the original config file.")
parser.add_argument(
"--pytorch_dump_folder_path", required=True, type=str, help="Path to the output PyTorch model directory."
)
_snake_case = parser.parse_args()
convert_mobilevitva_checkpoint(
args.task, args.orig_checkpoint_path, args.orig_config_path, args.pytorch_dump_folder_path
)
| 26 | 1 |
import unittest
from transformers import BertGenerationTokenizer
from transformers.testing_utils import get_tests_dir, require_sentencepiece, require_torch, slow
from transformers.utils import cached_property
from ...test_tokenization_common import TokenizerTesterMixin
_snake_case = "▁"
_snake_case = get_tests_dir("fixtures/test_sentencepiece.model")
@require_sentencepiece
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = BertGenerationTokenizer
_a = False
_a = True
def a__ ( self ) -> str:
super().setUp()
_A : Tuple = BertGenerationTokenizer(_a , keep_accents=_a )
tokenizer.save_pretrained(self.tmpdirname )
def a__ ( self ) -> Union[str, Any]:
_A : List[Any] = """<s>"""
_A : Optional[int] = 1
self.assertEqual(self.get_tokenizer()._convert_token_to_id(_a ) , _a )
self.assertEqual(self.get_tokenizer()._convert_id_to_token(_a ) , _a )
def a__ ( self ) -> List[Any]:
_A : List[str] = list(self.get_tokenizer().get_vocab().keys() )
self.assertEqual(vocab_keys[0] , """<unk>""" )
self.assertEqual(vocab_keys[1] , """<s>""" )
self.assertEqual(vocab_keys[-1] , """<pad>""" )
self.assertEqual(len(_a ) , 1002 )
def a__ ( self ) -> Union[str, Any]:
self.assertEqual(self.get_tokenizer().vocab_size , 1000 )
def a__ ( self ) -> Union[str, Any]:
_A : str = BertGenerationTokenizer(_a , keep_accents=_a )
_A : Tuple = tokenizer.tokenize("""This is a test""" )
self.assertListEqual(_a , ["""▁This""", """▁is""", """▁a""", """▁t""", """est"""] )
self.assertListEqual(
tokenizer.convert_tokens_to_ids(_a ) , [285, 46, 10, 170, 382] , )
_A : Union[str, Any] = tokenizer.tokenize("""I was born in 92000, and this is falsé.""" )
self.assertListEqual(
_a , [
SPIECE_UNDERLINE + """I""",
SPIECE_UNDERLINE + """was""",
SPIECE_UNDERLINE + """b""",
"""or""",
"""n""",
SPIECE_UNDERLINE + """in""",
SPIECE_UNDERLINE + """""",
"""9""",
"""2""",
"""0""",
"""0""",
"""0""",
""",""",
SPIECE_UNDERLINE + """and""",
SPIECE_UNDERLINE + """this""",
SPIECE_UNDERLINE + """is""",
SPIECE_UNDERLINE + """f""",
"""al""",
"""s""",
"""é""",
""".""",
] , )
_A : List[str] = tokenizer.convert_tokens_to_ids(_a )
self.assertListEqual(
_a , [8, 21, 84, 55, 24, 19, 7, 0, 602, 347, 347, 347, 3, 12, 66, 46, 72, 80, 6, 0, 4] , )
_A : Union[str, Any] = tokenizer.convert_ids_to_tokens(_a )
self.assertListEqual(
_a , [
SPIECE_UNDERLINE + """I""",
SPIECE_UNDERLINE + """was""",
SPIECE_UNDERLINE + """b""",
"""or""",
"""n""",
SPIECE_UNDERLINE + """in""",
SPIECE_UNDERLINE + """""",
"""<unk>""",
"""2""",
"""0""",
"""0""",
"""0""",
""",""",
SPIECE_UNDERLINE + """and""",
SPIECE_UNDERLINE + """this""",
SPIECE_UNDERLINE + """is""",
SPIECE_UNDERLINE + """f""",
"""al""",
"""s""",
"""<unk>""",
""".""",
] , )
@cached_property
def a__ ( self ) -> Tuple:
return BertGenerationTokenizer.from_pretrained("""google/bert_for_seq_generation_L-24_bbc_encoder""" )
@slow
def a__ ( self ) -> List[Any]:
_A : List[Any] = """Hello World!"""
_A : List[str] = [1_8536, 2260, 101]
self.assertListEqual(_a , self.big_tokenizer.encode(_a ) )
@slow
def a__ ( self ) -> List[str]:
_A : str = (
"""This is a very long text with a lot of weird characters, such as: . , ~ ? ( ) \" [ ] ! : - . Also we will"""
""" add words that should not exsist and be tokenized to <unk>, such as saoneuhaoesuth"""
)
_A : List[Any] = [
871,
419,
358,
946,
991,
2521,
452,
358,
1357,
387,
7751,
3536,
112,
985,
456,
126,
865,
938,
5400,
5734,
458,
1368,
467,
786,
2462,
5246,
1159,
633,
865,
4519,
457,
582,
852,
2557,
427,
916,
508,
405,
3_4324,
497,
391,
408,
1_1342,
1244,
385,
100,
938,
985,
456,
574,
362,
1_2597,
3200,
3129,
1172,
]
self.assertListEqual(_a , self.big_tokenizer.encode(_a ) )
@require_torch
@slow
def a__ ( self ) -> Optional[Any]:
import torch
from transformers import BertGenerationConfig, BertGenerationEncoder
# Build sequence
_A : Optional[int] = list(self.big_tokenizer.get_vocab().keys() )[:10]
_A : str = """ """.join(_a )
_A : int = self.big_tokenizer.encode_plus(_a , return_tensors="""pt""" , return_token_type_ids=_a )
_A : Any = self.big_tokenizer.batch_encode_plus(
[sequence + """ """ + sequence] , return_tensors="""pt""" , return_token_type_ids=_a )
_A : Optional[int] = BertGenerationConfig()
_A : int = BertGenerationEncoder(_a )
assert model.get_input_embeddings().weight.shape[0] >= self.big_tokenizer.vocab_size
with torch.no_grad():
model(**_a )
model(**_a )
@slow
def a__ ( self ) -> Optional[Any]:
# fmt: off
_A : Any = {"""input_ids""": [[3_9286, 458, 3_6335, 2001, 456, 1_3073, 1_3266, 455, 113, 7746, 1741, 1_1157, 391, 1_3073, 1_3266, 455, 113, 3967, 3_5412, 113, 4936, 109, 3870, 2377, 113, 3_0084, 4_5720, 458, 134, 1_7496, 112, 503, 1_1672, 113, 118, 112, 5665, 1_3347, 3_8687, 112, 1496, 3_1389, 112, 3268, 4_7264, 134, 962, 112, 1_6377, 8035, 2_3130, 430, 1_2169, 1_5518, 2_8592, 458, 146, 4_1697, 109, 391, 1_2169, 1_5518, 1_6689, 458, 146, 4_1358, 109, 452, 726, 4034, 111, 763, 3_5412, 5082, 388, 1903, 111, 9051, 391, 2870, 4_8918, 1900, 1123, 550, 998, 112, 9586, 1_5985, 455, 391, 410, 2_2955, 3_7636, 114], [448, 1_7496, 419, 3663, 385, 763, 113, 2_7533, 2870, 3283, 1_3043, 1639, 2_4713, 523, 656, 2_4013, 1_8550, 2521, 517, 2_7014, 2_1244, 420, 1212, 1465, 391, 927, 4833, 388, 578, 1_1786, 114, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [484, 2169, 7687, 2_1932, 1_8146, 726, 363, 1_7032, 3391, 114, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """attention_mask""": [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # noqa: E501
# fmt: on
self.tokenizer_integration_test_util(
expected_encoding=_a , model_name="""google/bert_for_seq_generation_L-24_bbc_encoder""" , revision="""c817d1fd1be2ffa69431227a1fe320544943d4db""" , )
| 26 |
import torch
from diffusers import DPMSolverSDEScheduler
from diffusers.utils import torch_device
from diffusers.utils.testing_utils import require_torchsde
from .test_schedulers import SchedulerCommonTest
@require_torchsde
class lowercase ( UpperCamelCase__ ):
_a = (DPMSolverSDEScheduler,)
_a = 1_0
def a__ ( self , **_a ) -> Optional[Any]:
_A : str = {
"""num_train_timesteps""": 1100,
"""beta_start""": 0.0001,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
"""noise_sampler_seed""": 0,
}
config.update(**_a )
return config
def a__ ( self ) -> Tuple:
for timesteps in [10, 50, 100, 1000]:
self.check_over_configs(num_train_timesteps=_a )
def a__ ( self ) -> Optional[int]:
for beta_start, beta_end in zip([0.00001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ):
self.check_over_configs(beta_start=_a , beta_end=_a )
def a__ ( self ) -> Any:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=_a )
def a__ ( self ) -> Optional[int]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=_a )
def a__ ( self ) -> Optional[int]:
_A : Any = self.scheduler_classes[0]
_A : List[str] = self.get_scheduler_config()
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Dict = self.dummy_model()
_A : Any = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Dict = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : str = model(_a , _a )
_A : List[Any] = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Dict = torch.sum(torch.abs(_a ) )
_A : Dict = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.47821044921875 ) < 1e-2
assert abs(result_mean.item() - 0.2178705964565277 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59352111816406 ) < 1e-2
assert abs(result_mean.item() - 0.22342906892299652 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Optional[Any]:
_A : Dict = self.scheduler_classes[0]
_A : Optional[int] = self.get_scheduler_config(prediction_type="""v_prediction""" )
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Tuple = self.dummy_model()
_A : int = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Tuple = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : int = scheduler.scale_model_input(_a , _a )
_A : Tuple = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Optional[Any] = torch.sum(torch.abs(_a ) )
_A : List[Any] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 124.77149200439453 ) < 1e-2
assert abs(result_mean.item() - 0.16226289014816284 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 128.1663360595703 ) < 1e-2
assert abs(result_mean.item() - 0.16688326001167297 ) < 1e-3
else:
assert abs(result_sum.item() - 119.8487548828125 ) < 1e-2
assert abs(result_mean.item() - 0.1560530662536621 ) < 1e-3
def a__ ( self ) -> List[str]:
_A : Union[str, Any] = self.scheduler_classes[0]
_A : List[Any] = self.get_scheduler_config()
_A : List[str] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Union[str, Any] = self.dummy_model()
_A : Optional[Any] = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
_A : int = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Dict = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : str = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.46957397460938 ) < 1e-2
assert abs(result_mean.item() - 0.21805934607982635 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59353637695312 ) < 1e-2
assert abs(result_mean.item() - 0.22342908382415771 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Union[str, Any]:
_A : List[Any] = self.scheduler_classes[0]
_A : Optional[Any] = self.get_scheduler_config()
_A : int = scheduler_class(**_a , use_karras_sigmas=_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Optional[Any] = self.dummy_model()
_A : Dict = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
_A : str = sample.to(_a )
for t in scheduler.timesteps:
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : List[str] = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : List[str] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 176.66974135742188 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 177.63653564453125 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
else:
assert abs(result_sum.item() - 170.3135223388672 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
| 26 | 1 |
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_flax_available,
is_tf_available,
is_tokenizers_available,
is_torch_available,
)
_snake_case = {
"configuration_roformer": ["ROFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP", "RoFormerConfig", "RoFormerOnnxConfig"],
"tokenization_roformer": ["RoFormerTokenizer"],
}
try:
if not is_tokenizers_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = ["RoFormerTokenizerFast"]
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"ROFORMER_PRETRAINED_MODEL_ARCHIVE_LIST",
"RoFormerForCausalLM",
"RoFormerForMaskedLM",
"RoFormerForMultipleChoice",
"RoFormerForQuestionAnswering",
"RoFormerForSequenceClassification",
"RoFormerForTokenClassification",
"RoFormerLayer",
"RoFormerModel",
"RoFormerPreTrainedModel",
"load_tf_weights_in_roformer",
]
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"TF_ROFORMER_PRETRAINED_MODEL_ARCHIVE_LIST",
"TFRoFormerForCausalLM",
"TFRoFormerForMaskedLM",
"TFRoFormerForMultipleChoice",
"TFRoFormerForQuestionAnswering",
"TFRoFormerForSequenceClassification",
"TFRoFormerForTokenClassification",
"TFRoFormerLayer",
"TFRoFormerModel",
"TFRoFormerPreTrainedModel",
]
try:
if not is_flax_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"FLAX_ROFORMER_PRETRAINED_MODEL_ARCHIVE_LIST",
"FlaxRoFormerForMaskedLM",
"FlaxRoFormerForMultipleChoice",
"FlaxRoFormerForQuestionAnswering",
"FlaxRoFormerForSequenceClassification",
"FlaxRoFormerForTokenClassification",
"FlaxRoFormerModel",
"FlaxRoFormerPreTrainedModel",
]
if TYPE_CHECKING:
from .configuration_roformer import ROFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP, RoFormerConfig, RoFormerOnnxConfig
from .tokenization_roformer import RoFormerTokenizer
try:
if not is_tokenizers_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .tokenization_roformer_fast import RoFormerTokenizerFast
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_roformer import (
ROFORMER_PRETRAINED_MODEL_ARCHIVE_LIST,
RoFormerForCausalLM,
RoFormerForMaskedLM,
RoFormerForMultipleChoice,
RoFormerForQuestionAnswering,
RoFormerForSequenceClassification,
RoFormerForTokenClassification,
RoFormerLayer,
RoFormerModel,
RoFormerPreTrainedModel,
load_tf_weights_in_roformer,
)
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_tf_roformer import (
TF_ROFORMER_PRETRAINED_MODEL_ARCHIVE_LIST,
TFRoFormerForCausalLM,
TFRoFormerForMaskedLM,
TFRoFormerForMultipleChoice,
TFRoFormerForQuestionAnswering,
TFRoFormerForSequenceClassification,
TFRoFormerForTokenClassification,
TFRoFormerLayer,
TFRoFormerModel,
TFRoFormerPreTrainedModel,
)
try:
if not is_flax_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_flax_roformer import (
FLAX_ROFORMER_PRETRAINED_MODEL_ARCHIVE_LIST,
FlaxRoFormerForMaskedLM,
FlaxRoFormerForMultipleChoice,
FlaxRoFormerForQuestionAnswering,
FlaxRoFormerForSequenceClassification,
FlaxRoFormerForTokenClassification,
FlaxRoFormerModel,
FlaxRoFormerPreTrainedModel,
)
else:
import sys
_snake_case = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
| 26 |
# DISCLAIMER: This file is strongly influenced by https://github.com/yang-song/score_sde_pytorch
import math
from typing import Union
import torch
from ..configuration_utils import ConfigMixin, register_to_config
from ..utils import randn_tensor
from .scheduling_utils import SchedulerMixin
class lowercase ( UpperCamelCase__,UpperCamelCase__ ):
_a = 1
@register_to_config
def __init__( self , _a=2000 , _a=0.1 , _a=20 , _a=1e-3 ) -> List[Any]:
_A : Dict = None
_A : List[Any] = None
_A : Dict = None
def a__ ( self , _a , _a = None ) -> Union[str, Any]:
_A : Union[str, Any] = torch.linspace(1 , self.config.sampling_eps , _a , device=_a )
def a__ ( self , _a , _a , _a , _a=None ) -> Dict:
if self.timesteps is None:
raise ValueError(
"""`self.timesteps` is not set, you need to run 'set_timesteps' after creating the scheduler""" )
# TODO(Patrick) better comments + non-PyTorch
# postprocess model score
_A : Any = (
-0.25 * t**2 * (self.config.beta_max - self.config.beta_min) - 0.5 * t * self.config.beta_min
)
_A : List[Any] = torch.sqrt(1.0 - torch.exp(2.0 * log_mean_coeff ) )
_A : List[str] = std.flatten()
while len(std.shape ) < len(score.shape ):
_A : List[Any] = std.unsqueeze(-1 )
_A : int = -score / std
# compute
_A : Tuple = -1.0 / len(self.timesteps )
_A : str = self.config.beta_min + t * (self.config.beta_max - self.config.beta_min)
_A : List[str] = beta_t.flatten()
while len(beta_t.shape ) < len(x.shape ):
_A : Union[str, Any] = beta_t.unsqueeze(-1 )
_A : Tuple = -0.5 * beta_t * x
_A : Tuple = torch.sqrt(_a )
_A : Dict = drift - diffusion**2 * score
_A : Dict = x + drift * dt
# add noise
_A : Any = randn_tensor(x.shape , layout=x.layout , generator=_a , device=x.device , dtype=x.dtype )
_A : str = x_mean + diffusion * math.sqrt(-dt ) * noise
return x, x_mean
def __len__( self ) -> Optional[Any]:
return self.config.num_train_timesteps
| 26 | 1 |
def lowerCAmelCase_ ( snake_case_ ):
_A : Optional[int] = set()
# edges = list of graph's edges
_A : int = get_edges(snake_case_ )
# While there are still elements in edges list, take an arbitrary edge
# (from_node, to_node) and add his extremity to chosen_vertices and then
# remove all arcs adjacent to the from_node and to_node
while edges:
_A , _A : Any = edges.pop()
chosen_vertices.add(snake_case_ )
chosen_vertices.add(snake_case_ )
for edge in edges.copy():
if from_node in edge or to_node in edge:
edges.discard(snake_case_ )
return chosen_vertices
def lowerCAmelCase_ ( snake_case_ ):
_A : Optional[Any] = set()
for from_node, to_nodes in graph.items():
for to_node in to_nodes:
edges.add((from_node, to_node) )
return edges
if __name__ == "__main__":
import doctest
doctest.testmod()
# graph = {0: [1, 3], 1: [0, 3], 2: [0, 3, 4], 3: [0, 1, 2], 4: [2, 3]}
# print(f"Matching vertex cover:\n{matching_min_vertex_cover(graph)}")
| 26 |
import os
from shutil import copyfile
from typing import List, Optional, Tuple
from ...tokenization_utils import AddedToken
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import is_sentencepiece_available, logging
if is_sentencepiece_available():
from .tokenization_fnet import FNetTokenizer
else:
_snake_case = None
_snake_case = logging.get_logger(__name__)
_snake_case = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.json"}
_snake_case = {
"vocab_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/spiece.model",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/spiece.model",
},
"tokenizer_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/tokenizer.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/tokenizer.json",
},
}
_snake_case = {
"google/fnet-base": 512,
"google/fnet-large": 512,
}
_snake_case = "▁"
class lowercase ( UpperCamelCase__ ):
_a = VOCAB_FILES_NAMES
_a = PRETRAINED_VOCAB_FILES_MAP
_a = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
_a = ["input_ids", "token_type_ids"]
_a = FNetTokenizer
def __init__( self , _a=None , _a=None , _a=False , _a=True , _a=True , _a="<unk>" , _a="[SEP]" , _a="<pad>" , _a="[CLS]" , _a="[MASK]" , **_a , ) -> Optional[int]:
# Mask token behave like a normal word, i.e. include the space before it and
# is included in the raw text, there should be a match in a non-normalized sentence.
_A : int = (
AddedToken(_a , lstrip=_a , rstrip=_a , normalized=_a )
if isinstance(_a , _a )
else mask_token
)
super().__init__(
_a , tokenizer_file=_a , do_lower_case=_a , remove_space=_a , keep_accents=_a , unk_token=_a , sep_token=_a , pad_token=_a , cls_token=_a , mask_token=_a , **_a , )
_A : Optional[int] = do_lower_case
_A : List[Any] = remove_space
_A : str = keep_accents
_A : int = vocab_file
_A : int = False if not self.vocab_file else True
def a__ ( self , _a , _a = None ) -> List[int]:
_A : str = [self.sep_token_id]
_A : Dict = [self.cls_token_id]
if token_ids_a is None:
return cls + token_ids_a + sep
return cls + token_ids_a + sep + token_ids_a + sep
def a__ ( self , _a , _a = None ) -> List[int]:
_A : Any = [self.sep_token_id]
_A : Any = [self.cls_token_id]
if token_ids_a is None:
return len(cls + token_ids_a + sep ) * [0]
return len(cls + token_ids_a + sep ) * [0] + len(token_ids_a + sep ) * [1]
def a__ ( self , _a , _a = None ) -> Tuple[str]:
if not os.path.isdir(_a ):
logger.error(F'''Vocabulary path ({save_directory}) should be a directory''' )
return
_A : List[str] = os.path.join(
_a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
if os.path.abspath(self.vocab_file ) != os.path.abspath(_a ):
copyfile(self.vocab_file , _a )
return (out_vocab_file,)
| 26 | 1 |
import itertools
import math
def lowerCAmelCase_ ( snake_case_ ):
if 1 < number < 4:
# 2 and 3 are primes
return True
elif number < 2 or number % 2 == 0 or number % 3 == 0:
# Negatives, 0, 1, all even numbers, all multiples of 3 are not primes
return False
# All primes number are in format of 6k +/- 1
for i in range(5,int(math.sqrt(snake_case_ ) + 1 ),6 ):
if number % i == 0 or number % (i + 2) == 0:
return False
return True
def lowerCAmelCase_ ( ):
_A : Dict = 2
while True:
if is_prime(snake_case_ ):
yield num
num += 1
def lowerCAmelCase_ ( snake_case_ = 10001 ):
return next(itertools.islice(prime_generator(),nth - 1,snake_case_ ) )
if __name__ == "__main__":
print(f"""{solution() = }""")
| 26 |
from math import asin, atan, cos, radians, sin, sqrt, tan
_snake_case = 6_3_7_8_1_3_7.0
_snake_case = 6_3_5_6_7_5_2.3_1_4_2_4_5
_snake_case = 6378137
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : Any = (AXIS_A - AXIS_B) / AXIS_A
_A : Optional[int] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : List[str] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : Optional[Any] = radians(snake_case_ )
_A : str = radians(snake_case_ )
# Equation
_A : Dict = sin((phi_a - phi_a) / 2 )
_A : List[str] = sin((lambda_a - lambda_a) / 2 )
# Square both values
sin_sq_phi *= sin_sq_phi
sin_sq_lambda *= sin_sq_lambda
_A : Optional[int] = sqrt(sin_sq_phi + (cos(snake_case_ ) * cos(snake_case_ ) * sin_sq_lambda) )
return 2 * RADIUS * asin(snake_case_ )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
import random
import unittest
import torch
from diffusers import IFImgaImgSuperResolutionPipeline
from diffusers.utils import floats_tensor
from diffusers.utils.import_utils import is_xformers_available
from diffusers.utils.testing_utils import skip_mps, torch_device
from ..pipeline_params import TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS, TEXT_GUIDED_IMAGE_VARIATION_PARAMS
from ..test_pipelines_common import PipelineTesterMixin
from . import IFPipelineTesterMixin
@skip_mps
class lowercase ( UpperCamelCase__,UpperCamelCase__,unittest.TestCase ):
_a = IFImgaImgSuperResolutionPipeline
_a = TEXT_GUIDED_IMAGE_VARIATION_PARAMS - {"width", "height"}
_a = TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS.union({"original_image"} )
_a = PipelineTesterMixin.required_optional_params - {"latents"}
def a__ ( self ) -> List[Any]:
return self._get_superresolution_dummy_components()
def a__ ( self , _a , _a=0 ) -> Any:
if str(_a ).startswith("""mps""" ):
_A : Tuple = torch.manual_seed(_a )
else:
_A : Tuple = torch.Generator(device=_a ).manual_seed(_a )
_A : List[Any] = floats_tensor((1, 3, 32, 32) , rng=random.Random(_a ) ).to(_a )
_A : Dict = floats_tensor((1, 3, 16, 16) , rng=random.Random(_a ) ).to(_a )
_A : Optional[int] = {
"""prompt""": """A painting of a squirrel eating a burger""",
"""image""": image,
"""original_image""": original_image,
"""generator""": generator,
"""num_inference_steps""": 2,
"""output_type""": """numpy""",
}
return inputs
@unittest.skipIf(
torch_device != """cuda""" or not is_xformers_available() , reason="""XFormers attention is only available with CUDA and `xformers` installed""" , )
def a__ ( self ) -> Optional[int]:
self._test_xformers_attention_forwardGenerator_pass(expected_max_diff=1e-3 )
def a__ ( self ) -> List[Any]:
self._test_save_load_optional_components()
@unittest.skipIf(torch_device != """cuda""" , reason="""float16 requires CUDA""" )
def a__ ( self ) -> Any:
# Due to non-determinism in save load of the hf-internal-testing/tiny-random-t5 text encoder
super().test_save_load_floataa(expected_max_diff=1e-1 )
def a__ ( self ) -> Union[str, Any]:
self._test_attention_slicing_forward_pass(expected_max_diff=1e-2 )
def a__ ( self ) -> List[Any]:
self._test_save_load_local()
def a__ ( self ) -> str:
self._test_inference_batch_single_identical(
expected_max_diff=1e-2 , )
| 26 |
from __future__ import absolute_import, division, print_function, unicode_literals
from torch import nn
from torch.nn import CrossEntropyLoss, MSELoss
from transformers import RobertaConfig
from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
from transformers.models.roberta.modeling_roberta import (
ROBERTA_INPUTS_DOCSTRING,
ROBERTA_START_DOCSTRING,
RobertaEmbeddings,
)
from .modeling_highway_bert import BertPreTrainedModel, DeeBertModel, HighwayException, entropy
@add_start_docstrings(
"The RoBERTa Model transformer with early exiting (DeeRoBERTa). ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> Optional[int]:
super().__init__(_a )
_A : Union[str, Any] = RobertaEmbeddings(_a )
self.init_weights()
@add_start_docstrings(
"RoBERTa Model (with early exiting - DeeRoBERTa) with a classifier on top,\n also takes care of multi-layer training. ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> str:
super().__init__(_a )
_A : Any = config.num_labels
_A : Dict = config.num_hidden_layers
_A : List[str] = DeeRobertaModel(_a )
_A : int = nn.Dropout(config.hidden_dropout_prob )
_A : int = nn.Linear(config.hidden_size , self.config.num_labels )
@add_start_docstrings_to_model_forward(_a )
def a__ ( self , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=-1 , _a=False , ) -> Any:
_A : Optional[int] = self.num_layers
try:
_A : List[str] = self.roberta(
_a , attention_mask=_a , token_type_ids=_a , position_ids=_a , head_mask=_a , inputs_embeds=_a , )
_A : List[str] = outputs[1]
_A : List[str] = self.dropout(_a )
_A : Optional[Any] = self.classifier(_a )
_A : List[Any] = (logits,) + outputs[2:] # add hidden states and attention if they are here
except HighwayException as e:
_A : List[Any] = e.message
_A : Optional[int] = e.exit_layer
_A : Optional[int] = outputs[0]
if not self.training:
_A : int = entropy(_a )
_A : int = []
_A : int = []
if labels is not None:
if self.num_labels == 1:
# We are doing regression
_A : Union[str, Any] = MSELoss()
_A : Tuple = loss_fct(logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Dict = loss_fct(logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
# work with highway exits
_A : Optional[Any] = []
for highway_exit in outputs[-1]:
_A : Tuple = highway_exit[0]
if not self.training:
highway_logits_all.append(_a )
highway_entropy.append(highway_exit[2] )
if self.num_labels == 1:
# We are doing regression
_A : List[str] = MSELoss()
_A : Optional[int] = loss_fct(highway_logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Tuple = loss_fct(highway_logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
highway_losses.append(_a )
if train_highway:
_A : Dict = (sum(highway_losses[:-1] ),) + outputs
# exclude the final highway, of course
else:
_A : int = (loss,) + outputs
if not self.training:
_A : Optional[Any] = outputs + ((original_entropy, highway_entropy), exit_layer)
if output_layer >= 0:
_A : Union[str, Any] = (
(outputs[0],) + (highway_logits_all[output_layer],) + outputs[2:]
) # use the highway of the last layer
return outputs # (loss), logits, (hidden_states), (attentions), entropy
| 26 | 1 |
from ...processing_utils import ProcessorMixin
class lowercase ( UpperCamelCase__ ):
_a = "WhisperFeatureExtractor"
_a = "WhisperTokenizer"
def __init__( self , _a , _a ) -> str:
super().__init__(_a , _a )
_A : Dict = self.feature_extractor
_A : List[Any] = False
def a__ ( self , _a=None , _a=None , _a=True ) -> List[str]:
return self.tokenizer.get_decoder_prompt_ids(task=_a , language=_a , no_timestamps=_a )
def __call__( self , *_a , **_a ) -> List[str]:
# For backward compatibility
if self._in_target_context_manager:
return self.current_processor(*_a , **_a )
_A : Union[str, Any] = kwargs.pop("""audio""" , _a )
_A : Tuple = kwargs.pop("""sampling_rate""" , _a )
_A : int = kwargs.pop("""text""" , _a )
if len(_a ) > 0:
_A : Dict = args[0]
_A : int = args[1:]
if audio is None and text is None:
raise ValueError("""You need to specify either an `audio` or `text` input to process.""" )
if audio is not None:
_A : Optional[int] = self.feature_extractor(_a , *_a , sampling_rate=_a , **_a )
if text is not None:
_A : int = self.tokenizer(_a , **_a )
if text is None:
return inputs
elif audio is None:
return encodings
else:
_A : List[Any] = encodings["""input_ids"""]
return inputs
def a__ ( self , *_a , **_a ) -> Any:
return self.tokenizer.batch_decode(*_a , **_a )
def a__ ( self , *_a , **_a ) -> int:
return self.tokenizer.decode(*_a , **_a )
def a__ ( self , _a , _a="np" ) -> Optional[int]:
return self.tokenizer.get_prompt_ids(_a , return_tensors=_a )
| 26 |
from collections import OrderedDict
from typing import Mapping
from ...configuration_utils import PretrainedConfig
from ...onnx import OnnxConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"facebook/xmod-base": "https://huggingface.co/facebook/xmod-base/resolve/main/config.json",
"facebook/xmod-large-prenorm": "https://huggingface.co/facebook/xmod-large-prenorm/resolve/main/config.json",
"facebook/xmod-base-13-125k": "https://huggingface.co/facebook/xmod-base-13-125k/resolve/main/config.json",
"facebook/xmod-base-30-125k": "https://huggingface.co/facebook/xmod-base-30-125k/resolve/main/config.json",
"facebook/xmod-base-30-195k": "https://huggingface.co/facebook/xmod-base-30-195k/resolve/main/config.json",
"facebook/xmod-base-60-125k": "https://huggingface.co/facebook/xmod-base-60-125k/resolve/main/config.json",
"facebook/xmod-base-60-265k": "https://huggingface.co/facebook/xmod-base-60-265k/resolve/main/config.json",
"facebook/xmod-base-75-125k": "https://huggingface.co/facebook/xmod-base-75-125k/resolve/main/config.json",
"facebook/xmod-base-75-269k": "https://huggingface.co/facebook/xmod-base-75-269k/resolve/main/config.json",
}
class lowercase ( UpperCamelCase__ ):
_a = "xmod"
def __init__( self , _a=3_0522 , _a=768 , _a=12 , _a=12 , _a=3072 , _a="gelu" , _a=0.1 , _a=0.1 , _a=512 , _a=2 , _a=0.02 , _a=1e-12 , _a=1 , _a=0 , _a=2 , _a="absolute" , _a=True , _a=None , _a=False , _a=2 , _a=False , _a=True , _a=True , _a=("en_XX",) , _a=None , **_a , ) -> str:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Tuple = vocab_size
_A : Union[str, Any] = hidden_size
_A : Dict = num_hidden_layers
_A : Dict = num_attention_heads
_A : List[Any] = hidden_act
_A : Optional[Any] = intermediate_size
_A : Any = hidden_dropout_prob
_A : str = attention_probs_dropout_prob
_A : Dict = max_position_embeddings
_A : Any = type_vocab_size
_A : List[Any] = initializer_range
_A : int = layer_norm_eps
_A : int = position_embedding_type
_A : Any = use_cache
_A : int = classifier_dropout
_A : int = pre_norm
_A : Optional[Any] = adapter_reduction_factor
_A : List[Any] = adapter_layer_norm
_A : Optional[int] = adapter_reuse_layer_norm
_A : Any = ln_before_adapter
_A : Union[str, Any] = list(_a )
_A : List[Any] = default_language
class lowercase ( UpperCamelCase__ ):
@property
def a__ ( self ) -> Mapping[str, Mapping[int, str]]:
if self.task == "multiple-choice":
_A : Dict = {0: """batch""", 1: """choice""", 2: """sequence"""}
else:
_A : Dict = {0: """batch""", 1: """sequence"""}
return OrderedDict(
[
("""input_ids""", dynamic_axis),
("""attention_mask""", dynamic_axis),
] )
| 26 | 1 |
import argparse
from copy import deepcopy
import numpy as np
from datasets import ClassLabel, DatasetDict, load_dataset
from evaluate import load
from transformers import (
AutoModelForSequenceClassification,
AutoTokenizer,
DataCollatorWithPadding,
Trainer,
TrainerCallback,
TrainingArguments,
set_seed,
)
def lowerCAmelCase_ ( ):
_A : Optional[int] = argparse.ArgumentParser()
parser.add_argument("""--model_ckpt""",type=snake_case_,default="""microsoft/unixcoder-base-nine""" )
parser.add_argument("""--num_epochs""",type=snake_case_,default=5 )
parser.add_argument("""--batch_size""",type=snake_case_,default=6 )
parser.add_argument("""--gradient_accumulation_steps""",type=snake_case_,default=1 )
parser.add_argument("""--freeze""",type=snake_case_,default=snake_case_ )
parser.add_argument("""--learning_rate""",type=snake_case_,default=5e-4 )
parser.add_argument("""--seed""",type=snake_case_,default=0 )
parser.add_argument("""--lr_scheduler_type""",type=snake_case_,default="""cosine""" )
parser.add_argument("""--num_warmup_steps""",type=snake_case_,default=10 )
parser.add_argument("""--weight_decay""",type=snake_case_,default=0.01 )
parser.add_argument("""--output_dir""",type=snake_case_,default="""./results""" )
return parser.parse_args()
_snake_case = load("accuracy")
def lowerCAmelCase_ ( snake_case_ ):
_A , _A : str = eval_pred
_A : Optional[Any] = np.argmax(snake_case_,axis=1 )
return metric.compute(predictions=snake_case_,references=snake_case_ )
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a ) -> None:
super().__init__()
_A : Dict = trainer
def a__ ( self , _a , _a , _a , **_a ) -> str:
if control.should_evaluate:
_A : Tuple = deepcopy(_a )
self._trainer.evaluate(eval_dataset=self._trainer.train_dataset , metric_key_prefix="""train""" )
return control_copy
def lowerCAmelCase_ ( ):
_A : List[Any] = get_args()
set_seed(args.seed )
_A : Optional[int] = load_dataset("""codeparrot/codecomplex""",split="""train""" )
_A : Optional[int] = dataset.train_test_split(test_size=0.2 )
_A : str = train_test["""test"""].train_test_split(test_size=0.5 )
_A : Optional[int] = DatasetDict(
{
"""train""": train_test["""train"""],
"""test""": test_validation["""train"""],
"""valid""": test_validation["""test"""],
} )
print("""Loading tokenizer and model""" )
_A : List[Any] = AutoTokenizer.from_pretrained(args.model_ckpt )
_A : List[str] = tokenizer.eos_token
_A : int = AutoModelForSequenceClassification.from_pretrained(args.model_ckpt,num_labels=7 )
_A : Union[str, Any] = model.config.eos_token_id
if args.freeze:
for param in model.roberta.parameters():
_A : int = False
_A : int = ClassLabel(num_classes=7,names=list(set(train_test_validation["""train"""]["""complexity"""] ) ) )
def tokenize(snake_case_ ):
_A : List[str] = tokenizer(example["""src"""],truncation=snake_case_,max_length=1024 )
_A : List[Any] = labels.straint(example["""complexity"""] )
return {
"input_ids": inputs["input_ids"],
"attention_mask": inputs["attention_mask"],
"label": label,
}
_A : Optional[Any] = train_test_validation.map(
snake_case_,batched=snake_case_,remove_columns=train_test_validation["""train"""].column_names,)
_A : List[str] = DataCollatorWithPadding(tokenizer=snake_case_ )
_A : int = TrainingArguments(
output_dir=args.output_dir,learning_rate=args.learning_rate,lr_scheduler_type=args.lr_scheduler_type,evaluation_strategy="""epoch""",save_strategy="""epoch""",logging_strategy="""epoch""",per_device_train_batch_size=args.batch_size,per_device_eval_batch_size=args.batch_size,num_train_epochs=args.num_epochs,gradient_accumulation_steps=args.gradient_accumulation_steps,weight_decay=0.01,metric_for_best_model="""accuracy""",run_name="""complexity-java""",report_to="""wandb""",)
_A : Dict = Trainer(
model=snake_case_,args=snake_case_,train_dataset=tokenized_datasets["""train"""],eval_dataset=tokenized_datasets["""valid"""],tokenizer=snake_case_,data_collator=snake_case_,compute_metrics=snake_case_,)
print("""Training...""" )
trainer.add_callback(CustomCallback(snake_case_ ) )
trainer.train()
if __name__ == "__main__":
main()
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
if n == 0:
return 0
_A : Tuple = float("""-inf""" )
for i in range(1,n + 1 ):
_A : str = max(
snake_case_,prices[i - 1] + naive_cut_rod_recursive(n - i,snake_case_ ) )
return max_revue
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
_A : Dict = [float("""-inf""" ) for _ in range(n + 1 )]
return _top_down_cut_rod_recursive(snake_case_,snake_case_,snake_case_ )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
if max_rev[n] >= 0:
return max_rev[n]
elif n == 0:
return 0
else:
_A : List[str] = float("""-inf""" )
for i in range(1,n + 1 ):
_A : Optional[Any] = max(
snake_case_,prices[i - 1] + _top_down_cut_rod_recursive(n - i,snake_case_,snake_case_ ),)
_A : Tuple = max_revenue
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
# length(max_rev) = n + 1, to accommodate for the revenue obtainable from a rod of
# length 0.
_A : List[Any] = [float("""-inf""" ) for _ in range(n + 1 )]
_A : Any = 0
for i in range(1,n + 1 ):
_A : Optional[Any] = max_rev[i]
for j in range(1,i + 1 ):
_A : int = max(snake_case_,prices[j - 1] + max_rev[i - j] )
_A : int = max_revenue_i
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if n < 0:
_A : Optional[Any] = f'''n must be greater than or equal to 0. Got n = {n}'''
raise ValueError(snake_case_ )
if n > len(snake_case_ ):
_A : Any = (
"""Each integral piece of rod must have a corresponding price. """
f'''Got n = {n} but length of prices = {len(snake_case_ )}'''
)
raise ValueError(snake_case_ )
def lowerCAmelCase_ ( ):
_A : Tuple = [6, 10, 12, 15, 20, 23]
_A : List[Any] = len(snake_case_ )
# the best revenue comes from cutting the rod into 6 pieces, each
# of length 1 resulting in a revenue of 6 * 6 = 36.
_A : Any = 36
_A : List[Any] = top_down_cut_rod(snake_case_,snake_case_ )
_A : List[Any] = bottom_up_cut_rod(snake_case_,snake_case_ )
_A : Dict = naive_cut_rod_recursive(snake_case_,snake_case_ )
assert expected_max_revenue == max_rev_top_down
assert max_rev_top_down == max_rev_bottom_up
assert max_rev_bottom_up == max_rev_naive
if __name__ == "__main__":
main()
| 26 | 1 |
import itertools
import json
import os
import unittest
from transformers import AddedToken, RobertaTokenizer, RobertaTokenizerFast
from transformers.models.roberta.tokenization_roberta import VOCAB_FILES_NAMES
from transformers.testing_utils import require_tokenizers, slow
from ...test_tokenization_common import TokenizerTesterMixin
@require_tokenizers
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = RobertaTokenizer
_a = RobertaTokenizerFast
_a = True
_a = {"cls_token": "<s>"}
def a__ ( self ) -> Dict:
super().setUp()
# Adapted from Sennrich et al. 2015 and https://github.com/rsennrich/subword-nmt
_A : Any = [
"""l""",
"""o""",
"""w""",
"""e""",
"""r""",
"""s""",
"""t""",
"""i""",
"""d""",
"""n""",
"""\u0120""",
"""\u0120l""",
"""\u0120n""",
"""\u0120lo""",
"""\u0120low""",
"""er""",
"""\u0120lowest""",
"""\u0120newer""",
"""\u0120wider""",
"""<unk>""",
]
_A : List[str] = dict(zip(_a , range(len(_a ) ) ) )
_A : Union[str, Any] = ["""#version: 0.2""", """\u0120 l""", """\u0120l o""", """\u0120lo w""", """e r""", """"""]
_A : Optional[int] = {"""unk_token""": """<unk>"""}
_A : Tuple = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES["""vocab_file"""] )
_A : int = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES["""merges_file"""] )
with open(self.vocab_file , """w""" , encoding="""utf-8""" ) as fp:
fp.write(json.dumps(_a ) + """\n""" )
with open(self.merges_file , """w""" , encoding="""utf-8""" ) as fp:
fp.write("""\n""".join(_a ) )
def a__ ( self , **_a ) -> List[Any]:
kwargs.update(self.special_tokens_map )
return self.tokenizer_class.from_pretrained(self.tmpdirname , **_a )
def a__ ( self , **_a ) -> str:
kwargs.update(self.special_tokens_map )
return RobertaTokenizerFast.from_pretrained(self.tmpdirname , **_a )
def a__ ( self , _a ) -> Dict:
_A : Optional[int] = """lower newer"""
_A : int = """lower newer"""
return input_text, output_text
def a__ ( self ) -> List[str]:
_A : Tuple = self.tokenizer_class(self.vocab_file , self.merges_file , **self.special_tokens_map )
_A : List[str] = """lower newer"""
_A : str = ["""l""", """o""", """w""", """er""", """\u0120""", """n""", """e""", """w""", """er"""]
_A : str = tokenizer.tokenize(_a ) # , add_prefix_space=True)
self.assertListEqual(_a , _a )
_A : Optional[Any] = tokens + [tokenizer.unk_token]
_A : Any = [0, 1, 2, 15, 10, 9, 3, 2, 15, 19]
self.assertListEqual(tokenizer.convert_tokens_to_ids(_a ) , _a )
def a__ ( self ) -> List[Any]:
_A : Optional[int] = self.get_tokenizer()
self.assertListEqual(tokenizer.encode("""Hello world!""" , add_special_tokens=_a ) , [0, 3_1414, 232, 328, 2] )
self.assertListEqual(
tokenizer.encode("""Hello world! cécé herlolip 418""" , add_special_tokens=_a ) , [0, 3_1414, 232, 328, 740, 1140, 1_2695, 69, 4_6078, 1588, 2] , )
@slow
def a__ ( self ) -> str:
_A : Union[str, Any] = self.tokenizer_class.from_pretrained("""roberta-base""" )
_A : Optional[int] = tokenizer.encode("""sequence builders""" , add_special_tokens=_a )
_A : Union[str, Any] = tokenizer.encode("""multi-sequence build""" , add_special_tokens=_a )
_A : Optional[Any] = tokenizer.encode(
"""sequence builders""" , add_special_tokens=_a , add_prefix_space=_a )
_A : List[str] = tokenizer.encode(
"""sequence builders""" , """multi-sequence build""" , add_special_tokens=_a , add_prefix_space=_a )
_A : List[str] = tokenizer.build_inputs_with_special_tokens(_a )
_A : Dict = tokenizer.build_inputs_with_special_tokens(_a , _a )
assert encoded_sentence == encoded_text_from_decode
assert encoded_pair == encoded_pair_from_decode
def a__ ( self ) -> List[Any]:
_A : Optional[int] = self.get_tokenizer()
_A : List[Any] = """Encode this sequence."""
_A : Any = tokenizer.byte_encoder[""" """.encode("""utf-8""" )[0]]
# Testing encoder arguments
_A : Dict = tokenizer.encode(_a , add_special_tokens=_a , add_prefix_space=_a )
_A : Dict = tokenizer.convert_ids_to_tokens(encoded[0] )[0]
self.assertNotEqual(_a , _a )
_A : int = tokenizer.encode(_a , add_special_tokens=_a , add_prefix_space=_a )
_A : Tuple = tokenizer.convert_ids_to_tokens(encoded[0] )[0]
self.assertEqual(_a , _a )
tokenizer.add_special_tokens({"""bos_token""": """<s>"""} )
_A : List[Any] = tokenizer.encode(_a , add_special_tokens=_a )
_A : int = tokenizer.convert_ids_to_tokens(encoded[1] )[0]
self.assertNotEqual(_a , _a )
# Testing spaces after special tokens
_A : Dict = """<mask>"""
tokenizer.add_special_tokens(
{"""mask_token""": AddedToken(_a , lstrip=_a , rstrip=_a )} ) # mask token has a left space
_A : Any = tokenizer.convert_tokens_to_ids(_a )
_A : Union[str, Any] = """Encode <mask> sequence"""
_A : Optional[Any] = """Encode <mask>sequence"""
_A : List[Any] = tokenizer.encode(_a )
_A : int = encoded.index(_a )
_A : Any = tokenizer.convert_ids_to_tokens(encoded[mask_loc + 1] )[0]
self.assertEqual(_a , _a )
_A : Optional[Any] = tokenizer.encode(_a )
_A : Optional[Any] = encoded.index(_a )
_A : List[str] = tokenizer.convert_ids_to_tokens(encoded[mask_loc + 1] )[0]
self.assertNotEqual(_a , _a )
def a__ ( self ) -> Optional[int]:
pass
def a__ ( self ) -> Tuple:
for tokenizer, pretrained_name, kwargs in self.tokenizers_list:
with self.subTest(F'''{tokenizer.__class__.__name__} ({pretrained_name})''' ):
_A : Any = self.rust_tokenizer_class.from_pretrained(_a , **_a )
_A : int = self.tokenizer_class.from_pretrained(_a , **_a )
_A : Union[str, Any] = """A, <mask> AllenNLP sentence."""
_A : List[Any] = tokenizer_r.encode_plus(_a , add_special_tokens=_a , return_token_type_ids=_a )
_A : Optional[int] = tokenizer_p.encode_plus(_a , add_special_tokens=_a , return_token_type_ids=_a )
# token_type_ids should put 0 everywhere
self.assertEqual(sum(tokens_r["""token_type_ids"""] ) , sum(tokens_p["""token_type_ids"""] ) )
# attention_mask should put 1 everywhere, so sum over length should be 1
self.assertEqual(
sum(tokens_r["""attention_mask"""] ) / len(tokens_r["""attention_mask"""] ) , sum(tokens_p["""attention_mask"""] ) / len(tokens_p["""attention_mask"""] ) , )
_A : str = tokenizer_r.convert_ids_to_tokens(tokens_r["""input_ids"""] )
_A : Optional[Any] = tokenizer_p.convert_ids_to_tokens(tokens_p["""input_ids"""] )
# Rust correctly handles the space before the mask while python doesnt
self.assertSequenceEqual(tokens_p["""input_ids"""] , [0, 250, 6, 5_0264, 3823, 487, 2_1992, 3645, 4, 2] )
self.assertSequenceEqual(tokens_r["""input_ids"""] , [0, 250, 6, 5_0264, 3823, 487, 2_1992, 3645, 4, 2] )
self.assertSequenceEqual(
_a , ["""<s>""", """A""", """,""", """<mask>""", """ĠAllen""", """N""", """LP""", """Ġsentence""", """.""", """</s>"""] )
self.assertSequenceEqual(
_a , ["""<s>""", """A""", """,""", """<mask>""", """ĠAllen""", """N""", """LP""", """Ġsentence""", """.""", """</s>"""] )
def a__ ( self ) -> Any:
for trim_offsets, add_prefix_space in itertools.product([True, False] , repeat=2 ):
_A : List[Any] = self.rust_tokenizer_class.from_pretrained(
self.tmpdirname , use_fast=_a , add_prefix_space=_a , trim_offsets=_a )
_A : Union[str, Any] = json.loads(tokenizer_r.backend_tokenizer.pre_tokenizer.__getstate__() )
_A : Union[str, Any] = json.loads(tokenizer_r.backend_tokenizer.post_processor.__getstate__() )
self.assertEqual(pre_tokenizer_state["""add_prefix_space"""] , _a )
self.assertEqual(post_processor_state["""add_prefix_space"""] , _a )
self.assertEqual(post_processor_state["""trim_offsets"""] , _a )
def a__ ( self ) -> Union[str, Any]:
# Test which aims to verify that the offsets are well adapted to the argument `add_prefix_space` and
# `trim_offsets`
for tokenizer, pretrained_name, kwargs in self.tokenizers_list:
with self.subTest(F'''{tokenizer.__class__.__name__} ({pretrained_name})''' ):
_A : Optional[int] = """hello""" # `hello` is a token in the vocabulary of `pretrained_name`
_A : Dict = F'''{text_of_1_token} {text_of_1_token}'''
_A : Tuple = self.rust_tokenizer_class.from_pretrained(
_a , use_fast=_a , add_prefix_space=_a , trim_offsets=_a )
_A : Optional[int] = tokenizer_r(_a , return_offsets_mapping=_a , add_special_tokens=_a )
self.assertEqual(encoding.offset_mapping[0] , (0, len(_a )) )
self.assertEqual(
encoding.offset_mapping[1] , (len(_a ) + 1, len(_a ) + 1 + len(_a )) , )
_A : List[str] = self.rust_tokenizer_class.from_pretrained(
_a , use_fast=_a , add_prefix_space=_a , trim_offsets=_a )
_A : str = tokenizer_r(_a , return_offsets_mapping=_a , add_special_tokens=_a )
self.assertEqual(encoding.offset_mapping[0] , (0, len(_a )) )
self.assertEqual(
encoding.offset_mapping[1] , (len(_a ) + 1, len(_a ) + 1 + len(_a )) , )
_A : Dict = self.rust_tokenizer_class.from_pretrained(
_a , use_fast=_a , add_prefix_space=_a , trim_offsets=_a )
_A : List[str] = tokenizer_r(_a , return_offsets_mapping=_a , add_special_tokens=_a )
self.assertEqual(encoding.offset_mapping[0] , (0, len(_a )) )
self.assertEqual(
encoding.offset_mapping[1] , (len(_a ), len(_a ) + 1 + len(_a )) , )
_A : int = self.rust_tokenizer_class.from_pretrained(
_a , use_fast=_a , add_prefix_space=_a , trim_offsets=_a )
_A : List[Any] = tokenizer_r(_a , return_offsets_mapping=_a , add_special_tokens=_a )
self.assertEqual(encoding.offset_mapping[0] , (0, len(_a )) )
self.assertEqual(
encoding.offset_mapping[1] , (len(_a ), len(_a ) + 1 + len(_a )) , )
_A : Dict = F''' {text}'''
# tokenizer_r = self.rust_tokenizer_class.from_pretrained(
# pretrained_name, use_fast=True, add_prefix_space=True, trim_offsets=True
# )
# encoding = tokenizer_r(text, return_offsets_mapping=True, add_special_tokens=False)
# self.assertEqual(encoding.offset_mapping[0], (1, 1 + len(text_of_1_token)))
# self.assertEqual(
# encoding.offset_mapping[1],
# (1 + len(text_of_1_token) + 1, 1 + len(text_of_1_token) + 1 + len(text_of_1_token)),
# )
_A : Optional[Any] = self.rust_tokenizer_class.from_pretrained(
_a , use_fast=_a , add_prefix_space=_a , trim_offsets=_a )
_A : Union[str, Any] = tokenizer_r(_a , return_offsets_mapping=_a , add_special_tokens=_a )
self.assertEqual(encoding.offset_mapping[0] , (1, 1 + len(_a )) )
self.assertEqual(
encoding.offset_mapping[1] , (1 + len(_a ) + 1, 1 + len(_a ) + 1 + len(_a )) , )
_A : Optional[int] = self.rust_tokenizer_class.from_pretrained(
_a , use_fast=_a , add_prefix_space=_a , trim_offsets=_a )
_A : int = tokenizer_r(_a , return_offsets_mapping=_a , add_special_tokens=_a )
self.assertEqual(encoding.offset_mapping[0] , (0, 1 + len(_a )) )
self.assertEqual(
encoding.offset_mapping[1] , (1 + len(_a ), 1 + len(_a ) + 1 + len(_a )) , )
_A : Optional[int] = self.rust_tokenizer_class.from_pretrained(
_a , use_fast=_a , add_prefix_space=_a , trim_offsets=_a )
_A : Optional[Any] = tokenizer_r(_a , return_offsets_mapping=_a , add_special_tokens=_a )
self.assertEqual(encoding.offset_mapping[0] , (0, 1 + len(_a )) )
self.assertEqual(
encoding.offset_mapping[1] , (1 + len(_a ), 1 + len(_a ) + 1 + len(_a )) , )
| 26 |
import requests
from bsa import BeautifulSoup
def lowerCAmelCase_ ( snake_case_ = "AAPL" ):
_A : str = f'''https://in.finance.yahoo.com/quote/{symbol}?s={symbol}'''
_A : List[Any] = BeautifulSoup(requests.get(snake_case_ ).text,"""html.parser""" )
_A : Union[str, Any] = """My(6px) Pos(r) smartphone_Mt(6px)"""
return soup.find("""div""",class_=class_ ).find("""span""" ).text
if __name__ == "__main__":
for symbol in "AAPL AMZN IBM GOOG MSFT ORCL".split():
print(f"""Current {symbol:<4} stock price is {stock_price(symbol):>8}""")
| 26 | 1 |
import torch
from diffusers import DPMSolverSDEScheduler
from diffusers.utils import torch_device
from diffusers.utils.testing_utils import require_torchsde
from .test_schedulers import SchedulerCommonTest
@require_torchsde
class lowercase ( UpperCamelCase__ ):
_a = (DPMSolverSDEScheduler,)
_a = 1_0
def a__ ( self , **_a ) -> Optional[Any]:
_A : str = {
"""num_train_timesteps""": 1100,
"""beta_start""": 0.0001,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
"""noise_sampler_seed""": 0,
}
config.update(**_a )
return config
def a__ ( self ) -> Tuple:
for timesteps in [10, 50, 100, 1000]:
self.check_over_configs(num_train_timesteps=_a )
def a__ ( self ) -> Optional[int]:
for beta_start, beta_end in zip([0.00001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ):
self.check_over_configs(beta_start=_a , beta_end=_a )
def a__ ( self ) -> Any:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=_a )
def a__ ( self ) -> Optional[int]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=_a )
def a__ ( self ) -> Optional[int]:
_A : Any = self.scheduler_classes[0]
_A : List[str] = self.get_scheduler_config()
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Dict = self.dummy_model()
_A : Any = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Dict = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : str = model(_a , _a )
_A : List[Any] = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Dict = torch.sum(torch.abs(_a ) )
_A : Dict = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.47821044921875 ) < 1e-2
assert abs(result_mean.item() - 0.2178705964565277 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59352111816406 ) < 1e-2
assert abs(result_mean.item() - 0.22342906892299652 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Optional[Any]:
_A : Dict = self.scheduler_classes[0]
_A : Optional[int] = self.get_scheduler_config(prediction_type="""v_prediction""" )
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Tuple = self.dummy_model()
_A : int = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Tuple = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : int = scheduler.scale_model_input(_a , _a )
_A : Tuple = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Optional[Any] = torch.sum(torch.abs(_a ) )
_A : List[Any] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 124.77149200439453 ) < 1e-2
assert abs(result_mean.item() - 0.16226289014816284 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 128.1663360595703 ) < 1e-2
assert abs(result_mean.item() - 0.16688326001167297 ) < 1e-3
else:
assert abs(result_sum.item() - 119.8487548828125 ) < 1e-2
assert abs(result_mean.item() - 0.1560530662536621 ) < 1e-3
def a__ ( self ) -> List[str]:
_A : Union[str, Any] = self.scheduler_classes[0]
_A : List[Any] = self.get_scheduler_config()
_A : List[str] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Union[str, Any] = self.dummy_model()
_A : Optional[Any] = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
_A : int = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Dict = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : str = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.46957397460938 ) < 1e-2
assert abs(result_mean.item() - 0.21805934607982635 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59353637695312 ) < 1e-2
assert abs(result_mean.item() - 0.22342908382415771 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Union[str, Any]:
_A : List[Any] = self.scheduler_classes[0]
_A : Optional[Any] = self.get_scheduler_config()
_A : int = scheduler_class(**_a , use_karras_sigmas=_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Optional[Any] = self.dummy_model()
_A : Dict = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
_A : str = sample.to(_a )
for t in scheduler.timesteps:
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : List[str] = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : List[str] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 176.66974135742188 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 177.63653564453125 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
else:
assert abs(result_sum.item() - 170.3135223388672 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
| 26 |
import unittest
from transformers import (
MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
TextaTextGenerationPipeline,
pipeline,
)
from transformers.testing_utils import is_pipeline_test, require_tf, require_torch
from transformers.utils import is_torch_available
from .test_pipelines_common import ANY
if is_torch_available():
import torch
@is_pipeline_test
class lowercase ( unittest.TestCase ):
_a = MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING
_a = TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING
def a__ ( self , _a , _a , _a ) -> int:
_A : str = TextaTextGenerationPipeline(model=_a , tokenizer=_a )
return generator, ["Something to write", "Something else"]
def a__ ( self , _a , _a ) -> Dict:
_A : Any = generator("""Something there""" )
self.assertEqual(_a , [{"""generated_text""": ANY(_a )}] )
# These are encoder decoder, they don't just append to incoming string
self.assertFalse(outputs[0]["""generated_text"""].startswith("""Something there""" ) )
_A : List[Any] = generator(["""This is great !""", """Something else"""] , num_return_sequences=2 , do_sample=_a )
self.assertEqual(
_a , [
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
] , )
_A : Optional[int] = generator(
["""This is great !""", """Something else"""] , num_return_sequences=2 , batch_size=2 , do_sample=_a )
self.assertEqual(
_a , [
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
[{"""generated_text""": ANY(_a )}, {"""generated_text""": ANY(_a )}],
] , )
with self.assertRaises(_a ):
generator(4 )
@require_torch
def a__ ( self ) -> List[str]:
_A : Any = pipeline("""text2text-generation""" , model="""patrickvonplaten/t5-tiny-random""" , framework="""pt""" )
# do_sample=False necessary for reproducibility
_A : Dict = generator("""Something there""" , do_sample=_a )
self.assertEqual(_a , [{"""generated_text""": """"""}] )
_A : Any = 3
_A : Any = generator(
"""Something there""" , num_return_sequences=_a , num_beams=_a , )
_A : Optional[int] = [
{"""generated_text""": """Beide Beide Beide Beide Beide Beide Beide Beide Beide"""},
{"""generated_text""": """Beide Beide Beide Beide Beide Beide Beide Beide"""},
{"""generated_text""": """"""},
]
self.assertEqual(_a , _a )
_A : Dict = generator("""This is a test""" , do_sample=_a , num_return_sequences=2 , return_tensors=_a )
self.assertEqual(
_a , [
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
] , )
_A : Dict = generator.model.config.eos_token_id
_A : List[str] = """<pad>"""
_A : Dict = generator(
["""This is a test""", """This is a second test"""] , do_sample=_a , num_return_sequences=2 , batch_size=2 , return_tensors=_a , )
self.assertEqual(
_a , [
[
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
],
[
{"""generated_token_ids""": ANY(torch.Tensor )},
{"""generated_token_ids""": ANY(torch.Tensor )},
],
] , )
@require_tf
def a__ ( self ) -> int:
_A : Optional[Any] = pipeline("""text2text-generation""" , model="""patrickvonplaten/t5-tiny-random""" , framework="""tf""" )
# do_sample=False necessary for reproducibility
_A : str = generator("""Something there""" , do_sample=_a )
self.assertEqual(_a , [{"""generated_text""": """"""}] )
| 26 | 1 |
import json
import os
import shutil
import tempfile
import unittest
from multiprocessing import get_context
from pathlib import Path
import datasets
import numpy as np
from datasets import load_dataset
from parameterized import parameterized
from transformers import AutoProcessor
from transformers.models.wavaveca import WavaVecaCTCTokenizer, WavaVecaFeatureExtractor
from transformers.models.wavaveca.tokenization_wavaveca import VOCAB_FILES_NAMES
from transformers.testing_utils import require_pyctcdecode, require_torch, require_torchaudio, slow
from transformers.utils import FEATURE_EXTRACTOR_NAME, is_pyctcdecode_available, is_torch_available
from ..wavaveca.test_feature_extraction_wavaveca import floats_list
if is_pyctcdecode_available():
from huggingface_hub import snapshot_download
from pyctcdecode import BeamSearchDecoderCTC
from transformers.models.wavaveca_with_lm import WavaVecaProcessorWithLM
from transformers.models.wavaveca_with_lm.processing_wavaveca_with_lm import WavaVecaDecoderWithLMOutput
if is_torch_available():
from transformers import WavaVecaForCTC
@require_pyctcdecode
class lowercase ( unittest.TestCase ):
def a__ ( self ) -> str:
_A : Any = """| <pad> <unk> <s> </s> a b c d e f g h i j k""".split()
_A : Tuple = dict(zip(_a , range(len(_a ) ) ) )
_A : List[str] = {
"""unk_token""": """<unk>""",
"""bos_token""": """<s>""",
"""eos_token""": """</s>""",
}
_A : Optional[Any] = {
"""feature_size""": 1,
"""padding_value""": 0.0,
"""sampling_rate""": 1_6000,
"""return_attention_mask""": False,
"""do_normalize""": True,
}
_A : Union[str, Any] = tempfile.mkdtemp()
_A : Union[str, Any] = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES["""vocab_file"""] )
_A : Optional[Any] = os.path.join(self.tmpdirname , _a )
with open(self.vocab_file , """w""" , encoding="""utf-8""" ) as fp:
fp.write(json.dumps(_a ) + """\n""" )
with open(self.feature_extraction_file , """w""" , encoding="""utf-8""" ) as fp:
fp.write(json.dumps(_a ) + """\n""" )
# load decoder from hub
_A : str = """hf-internal-testing/ngram-beam-search-decoder"""
def a__ ( self , **_a ) -> Union[str, Any]:
_A : str = self.add_kwargs_tokens_map.copy()
kwargs.update(_a )
return WavaVecaCTCTokenizer.from_pretrained(self.tmpdirname , **_a )
def a__ ( self , **_a ) -> Dict:
return WavaVecaFeatureExtractor.from_pretrained(self.tmpdirname , **_a )
def a__ ( self , **_a ) -> str:
return BeamSearchDecoderCTC.load_from_hf_hub(self.decoder_name , **_a )
def a__ ( self ) -> Any:
shutil.rmtree(self.tmpdirname )
def a__ ( self ) -> str:
_A : List[Any] = self.get_tokenizer()
_A : Optional[Any] = self.get_feature_extractor()
_A : List[str] = self.get_decoder()
_A : List[str] = WavaVecaProcessorWithLM(tokenizer=_a , feature_extractor=_a , decoder=_a )
processor.save_pretrained(self.tmpdirname )
_A : List[str] = WavaVecaProcessorWithLM.from_pretrained(self.tmpdirname )
# tokenizer
self.assertEqual(processor.tokenizer.get_vocab() , tokenizer.get_vocab() )
self.assertIsInstance(processor.tokenizer , _a )
# feature extractor
self.assertEqual(processor.feature_extractor.to_json_string() , feature_extractor.to_json_string() )
self.assertIsInstance(processor.feature_extractor , _a )
# decoder
self.assertEqual(processor.decoder._alphabet.labels , decoder._alphabet.labels )
self.assertEqual(
processor.decoder.model_container[decoder._model_key]._unigram_set , decoder.model_container[decoder._model_key]._unigram_set , )
self.assertIsInstance(processor.decoder , _a )
def a__ ( self ) -> Union[str, Any]:
_A : List[Any] = WavaVecaProcessorWithLM(
tokenizer=self.get_tokenizer() , feature_extractor=self.get_feature_extractor() , decoder=self.get_decoder() )
processor.save_pretrained(self.tmpdirname )
# make sure that error is thrown when decoder alphabet doesn't match
_A : int = WavaVecaProcessorWithLM.from_pretrained(
self.tmpdirname , alpha=5.0 , beta=3.0 , score_boundary=-7.0 , unk_score_offset=3 )
# decoder
self.assertEqual(processor.language_model.alpha , 5.0 )
self.assertEqual(processor.language_model.beta , 3.0 )
self.assertEqual(processor.language_model.score_boundary , -7.0 )
self.assertEqual(processor.language_model.unk_score_offset , 3 )
def a__ ( self ) -> Optional[Any]:
_A : Dict = self.get_tokenizer()
# add token to trigger raise
tokenizer.add_tokens(["""xx"""] )
with self.assertRaisesRegex(_a , """include""" ):
WavaVecaProcessorWithLM(
tokenizer=_a , feature_extractor=self.get_feature_extractor() , decoder=self.get_decoder() )
def a__ ( self ) -> List[str]:
_A : str = self.get_feature_extractor()
_A : Tuple = self.get_tokenizer()
_A : List[Any] = self.get_decoder()
_A : Any = WavaVecaProcessorWithLM(tokenizer=_a , feature_extractor=_a , decoder=_a )
_A : Tuple = floats_list((3, 1000) )
_A : Union[str, Any] = feature_extractor(_a , return_tensors="""np""" )
_A : Tuple = processor(_a , return_tensors="""np""" )
for key in input_feat_extract.keys():
self.assertAlmostEqual(input_feat_extract[key].sum() , input_processor[key].sum() , delta=1e-2 )
def a__ ( self ) -> str:
_A : Any = self.get_feature_extractor()
_A : List[str] = self.get_tokenizer()
_A : List[Any] = self.get_decoder()
_A : List[str] = WavaVecaProcessorWithLM(tokenizer=_a , feature_extractor=_a , decoder=_a )
_A : int = """This is a test string"""
_A : List[str] = processor(text=_a )
_A : Union[str, Any] = tokenizer(_a )
for key in encoded_tok.keys():
self.assertListEqual(encoded_tok[key] , encoded_processor[key] )
def a__ ( self , _a=(2, 10, 16) , _a=77 ) -> Optional[int]:
np.random.seed(_a )
return np.random.rand(*_a )
def a__ ( self ) -> List[str]:
_A : Dict = self.get_feature_extractor()
_A : List[Any] = self.get_tokenizer()
_A : Optional[int] = self.get_decoder()
_A : Optional[Any] = WavaVecaProcessorWithLM(tokenizer=_a , feature_extractor=_a , decoder=_a )
_A : List[Any] = self._get_dummy_logits(shape=(10, 16) , seed=13 )
_A : Dict = processor.decode(_a )
_A : Optional[Any] = decoder.decode_beams(_a )[0]
self.assertEqual(decoded_decoder[0] , decoded_processor.text )
self.assertEqual("""</s> <s> </s>""" , decoded_processor.text )
self.assertEqual(decoded_decoder[-2] , decoded_processor.logit_score )
self.assertEqual(decoded_decoder[-1] , decoded_processor.lm_score )
@parameterized.expand([[None], ["""fork"""], ["""spawn"""]] )
def a__ ( self , _a ) -> int:
_A : int = self.get_feature_extractor()
_A : Any = self.get_tokenizer()
_A : int = self.get_decoder()
_A : Tuple = WavaVecaProcessorWithLM(tokenizer=_a , feature_extractor=_a , decoder=_a )
_A : Union[str, Any] = self._get_dummy_logits()
# note: pool should be instantiated *after* Wav2Vec2ProcessorWithLM.
# otherwise, the LM won't be available to the pool's sub-processes.
# manual logic used to allow parameterized test for both pool=None and pool=Pool(...)
if pool_context is None:
_A : Any = processor.batch_decode(_a )
else:
with get_context(_a ).Pool() as pool:
_A : Optional[int] = processor.batch_decode(_a , _a )
_A : Optional[int] = list(_a )
with get_context("""fork""" ).Pool() as p:
_A : Tuple = decoder.decode_beams_batch(_a , _a )
_A , _A , _A : str = [], [], []
for beams in decoded_beams:
texts_decoder.append(beams[0][0] )
logit_scores_decoder.append(beams[0][-2] )
lm_scores_decoder.append(beams[0][-1] )
self.assertListEqual(_a , decoded_processor.text )
self.assertListEqual(["""<s> <s> </s>""", """<s> <s> <s>"""] , decoded_processor.text )
self.assertListEqual(_a , decoded_processor.logit_score )
self.assertListEqual(_a , decoded_processor.lm_score )
def a__ ( self ) -> Optional[Any]:
_A : Any = self.get_feature_extractor()
_A : str = self.get_tokenizer()
_A : int = self.get_decoder()
_A : Dict = WavaVecaProcessorWithLM(tokenizer=_a , feature_extractor=_a , decoder=_a )
_A : List[Any] = self._get_dummy_logits()
_A : Union[str, Any] = 15
_A : str = -20.0
_A : Optional[int] = -4.0
_A : Union[str, Any] = processor.batch_decode(
_a , beam_width=_a , beam_prune_logp=_a , token_min_logp=_a , )
_A : str = decoded_processor_out.text
_A : Dict = list(_a )
with get_context("""fork""" ).Pool() as pool:
_A : Union[str, Any] = decoder.decode_beams_batch(
_a , _a , beam_width=_a , beam_prune_logp=_a , token_min_logp=_a , )
_A : Optional[int] = [d[0][0] for d in decoded_decoder_out]
_A : int = [d[0][2] for d in decoded_decoder_out]
_A : Optional[int] = [d[0][3] for d in decoded_decoder_out]
self.assertListEqual(_a , _a )
self.assertListEqual(["""</s> <s> <s>""", """<s> <s> <s>"""] , _a )
self.assertTrue(np.array_equal(_a , decoded_processor_out.logit_score ) )
self.assertTrue(np.allclose([-20.054, -18.447] , _a , atol=1e-3 ) )
self.assertTrue(np.array_equal(_a , decoded_processor_out.lm_score ) )
self.assertTrue(np.allclose([-15.554, -13.9474] , _a , atol=1e-3 ) )
def a__ ( self ) -> str:
_A : Any = self.get_feature_extractor()
_A : Dict = self.get_tokenizer()
_A : Any = self.get_decoder()
_A : Any = WavaVecaProcessorWithLM(tokenizer=_a , feature_extractor=_a , decoder=_a )
_A : Dict = self._get_dummy_logits()
_A : Any = 2.0
_A : int = 5.0
_A : Union[str, Any] = -20.0
_A : str = True
_A : Optional[Any] = processor.batch_decode(
_a , alpha=_a , beta=_a , unk_score_offset=_a , lm_score_boundary=_a , )
_A : Union[str, Any] = decoded_processor_out.text
_A : Tuple = list(_a )
decoder.reset_params(
alpha=_a , beta=_a , unk_score_offset=_a , lm_score_boundary=_a , )
with get_context("""fork""" ).Pool() as pool:
_A : Optional[Any] = decoder.decode_beams_batch(
_a , _a , )
_A : Dict = [d[0][0] for d in decoded_decoder_out]
self.assertListEqual(_a , _a )
self.assertListEqual(["""<s> </s> <s> </s> </s>""", """</s> </s> <s> </s> </s>"""] , _a )
_A : List[Any] = processor.decoder.model_container[processor.decoder._model_key]
self.assertEqual(lm_model.alpha , 2.0 )
self.assertEqual(lm_model.beta , 5.0 )
self.assertEqual(lm_model.unk_score_offset , -20.0 )
self.assertEqual(lm_model.score_boundary , _a )
def a__ ( self ) -> Any:
_A : Tuple = WavaVecaProcessorWithLM.from_pretrained("""hf-internal-testing/processor_with_lm""" )
_A : Optional[int] = processor.decoder.model_container[processor.decoder._model_key]
_A : Optional[int] = Path(language_model._kenlm_model.path.decode("""utf-8""" ) ).parent.parent.absolute()
_A : Any = os.listdir(_a )
_A : int = ["""alphabet.json""", """language_model"""]
downloaded_decoder_files.sort()
expected_decoder_files.sort()
# test that only decoder relevant files from
# https://huggingface.co/hf-internal-testing/processor_with_lm/tree/main
# are downloaded and none of the rest (e.g. README.md, ...)
self.assertListEqual(_a , _a )
def a__ ( self ) -> Dict:
_A : int = snapshot_download("""hf-internal-testing/processor_with_lm""" )
_A : Optional[Any] = WavaVecaProcessorWithLM.from_pretrained(_a )
_A : Optional[int] = processor.decoder.model_container[processor.decoder._model_key]
_A : str = Path(language_model._kenlm_model.path.decode("""utf-8""" ) ).parent.parent.absolute()
_A : Optional[int] = os.listdir(_a )
_A : List[str] = os.listdir(_a )
local_decoder_files.sort()
expected_decoder_files.sort()
# test that both decoder form hub and local files in cache are the same
self.assertListEqual(_a , _a )
def a__ ( self ) -> Optional[Any]:
_A : List[str] = WavaVecaProcessorWithLM.from_pretrained("""hf-internal-testing/processor_with_lm""" )
_A : Union[str, Any] = AutoProcessor.from_pretrained("""hf-internal-testing/processor_with_lm""" )
_A : List[str] = floats_list((3, 1000) )
_A : Union[str, Any] = processor_wavaveca(_a , return_tensors="""np""" )
_A : Any = processor_auto(_a , return_tensors="""np""" )
for key in input_wavaveca.keys():
self.assertAlmostEqual(input_wavaveca[key].sum() , input_auto[key].sum() , delta=1e-2 )
_A : Tuple = self._get_dummy_logits()
_A : List[str] = processor_wavaveca.batch_decode(_a )
_A : Any = processor_auto.batch_decode(_a )
self.assertListEqual(decoded_wavaveca.text , decoded_auto.text )
def a__ ( self ) -> Dict:
_A : Dict = self.get_feature_extractor()
_A : Optional[int] = self.get_tokenizer()
_A : int = self.get_decoder()
_A : List[str] = WavaVecaProcessorWithLM(tokenizer=_a , feature_extractor=_a , decoder=_a )
self.assertListEqual(
processor.model_input_names , feature_extractor.model_input_names , msg="""`processor` and `feature_extractor` model input names do not match""" , )
@staticmethod
def a__ ( _a , _a ) -> int:
_A : List[Any] = [d[key] for d in offsets]
return retrieved_list
def a__ ( self ) -> int:
_A : Any = WavaVecaProcessorWithLM.from_pretrained("""hf-internal-testing/processor_with_lm""" )
_A : Optional[int] = self._get_dummy_logits()[0]
_A : str = processor.decode(_a , output_word_offsets=_a )
# check Wav2Vec2CTCTokenizerOutput keys for word
self.assertEqual(len(outputs.keys() ) , 4 )
self.assertTrue("""text""" in outputs )
self.assertTrue("""word_offsets""" in outputs )
self.assertTrue(isinstance(_a , _a ) )
self.assertEqual(""" """.join(self.get_from_offsets(outputs["""word_offsets"""] , """word""" ) ) , outputs.text )
self.assertListEqual(self.get_from_offsets(outputs["""word_offsets"""] , """word""" ) , ["""<s>""", """<s>""", """</s>"""] )
self.assertListEqual(self.get_from_offsets(outputs["""word_offsets"""] , """start_offset""" ) , [0, 2, 4] )
self.assertListEqual(self.get_from_offsets(outputs["""word_offsets"""] , """end_offset""" ) , [1, 3, 5] )
def a__ ( self ) -> int:
_A : Dict = WavaVecaProcessorWithLM.from_pretrained("""hf-internal-testing/processor_with_lm""" )
_A : Tuple = self._get_dummy_logits()
_A : int = processor.batch_decode(_a , output_word_offsets=_a )
# check Wav2Vec2CTCTokenizerOutput keys for word
self.assertEqual(len(outputs.keys() ) , 4 )
self.assertTrue("""text""" in outputs )
self.assertTrue("""word_offsets""" in outputs )
self.assertTrue(isinstance(_a , _a ) )
self.assertListEqual(
[""" """.join(self.get_from_offsets(_a , """word""" ) ) for o in outputs["""word_offsets"""]] , outputs.text )
self.assertListEqual(self.get_from_offsets(outputs["""word_offsets"""][0] , """word""" ) , ["""<s>""", """<s>""", """</s>"""] )
self.assertListEqual(self.get_from_offsets(outputs["""word_offsets"""][0] , """start_offset""" ) , [0, 2, 4] )
self.assertListEqual(self.get_from_offsets(outputs["""word_offsets"""][0] , """end_offset""" ) , [1, 3, 5] )
@slow
@require_torch
@require_torchaudio
def a__ ( self ) -> Tuple:
import torch
_A : Union[str, Any] = load_dataset("""common_voice""" , """en""" , split="""train""" , streaming=_a )
_A : List[Any] = ds.cast_column("""audio""" , datasets.Audio(sampling_rate=1_6000 ) )
_A : Optional[Any] = iter(_a )
_A : Any = next(_a )
_A : str = AutoProcessor.from_pretrained("""patrickvonplaten/wav2vec2-base-100h-with-lm""" )
_A : Optional[int] = WavaVecaForCTC.from_pretrained("""patrickvonplaten/wav2vec2-base-100h-with-lm""" )
# compare to filename `common_voice_en_100038.mp3` of dataset viewer on https://huggingface.co/datasets/common_voice/viewer/en/train
_A : Dict = processor(sample["""audio"""]["""array"""] , return_tensors="""pt""" ).input_values
with torch.no_grad():
_A : Dict = model(_a ).logits.cpu().numpy()
_A : Optional[int] = processor.decode(logits[0] , output_word_offsets=_a )
_A : List[str] = model.config.inputs_to_logits_ratio / processor.feature_extractor.sampling_rate
_A : List[Any] = [
{
"""start_time""": d["""start_offset"""] * time_offset,
"""end_time""": d["""end_offset"""] * time_offset,
"""word""": d["""word"""],
}
for d in output["""word_offsets"""]
]
_A : Optional[Any] = """WHY DOES MILISANDRA LOOK LIKE SHE WANTS TO CONSUME JOHN SNOW ON THE RIVER AT THE WALL"""
# output words
self.assertEqual(""" """.join(self.get_from_offsets(_a , """word""" ) ) , _a )
self.assertEqual(""" """.join(self.get_from_offsets(_a , """word""" ) ) , output.text )
# output times
_A : Optional[Any] = torch.tensor(self.get_from_offsets(_a , """start_time""" ) )
_A : int = torch.tensor(self.get_from_offsets(_a , """end_time""" ) )
# fmt: off
_A : Dict = torch.tensor([1.4199, 1.6599, 2.2599, 3.0, 3.24, 3.5999, 3.7999, 4.0999, 4.26, 4.94, 5.28, 5.6599, 5.78, 5.94, 6.32, 6.5399, 6.6599] )
_A : str = torch.tensor([1.5399, 1.8999, 2.9, 3.16, 3.5399, 3.72, 4.0199, 4.1799, 4.76, 5.1599, 5.5599, 5.6999, 5.86, 6.1999, 6.38, 6.6199, 6.94] )
# fmt: on
self.assertTrue(torch.allclose(_a , _a , atol=0.01 ) )
self.assertTrue(torch.allclose(_a , _a , atol=0.01 ) )
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
while b:
_A , _A : List[str] = b, a % b
return a
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
return a if b == 0 else euclidean_gcd_recursive(snake_case_,a % b )
def lowerCAmelCase_ ( ):
print(f'''euclidean_gcd(3, 5) = {euclidean_gcd(3,5 )}''' )
print(f'''euclidean_gcd(5, 3) = {euclidean_gcd(5,3 )}''' )
print(f'''euclidean_gcd(1, 3) = {euclidean_gcd(1,3 )}''' )
print(f'''euclidean_gcd(3, 6) = {euclidean_gcd(3,6 )}''' )
print(f'''euclidean_gcd(6, 3) = {euclidean_gcd(6,3 )}''' )
print(f'''euclidean_gcd_recursive(3, 5) = {euclidean_gcd_recursive(3,5 )}''' )
print(f'''euclidean_gcd_recursive(5, 3) = {euclidean_gcd_recursive(5,3 )}''' )
print(f'''euclidean_gcd_recursive(1, 3) = {euclidean_gcd_recursive(1,3 )}''' )
print(f'''euclidean_gcd_recursive(3, 6) = {euclidean_gcd_recursive(3,6 )}''' )
print(f'''euclidean_gcd_recursive(6, 3) = {euclidean_gcd_recursive(6,3 )}''' )
if __name__ == "__main__":
main()
| 26 | 1 |
import requests
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : List[str] = {"""Content-Type""": """application/json"""}
_A : str = requests.post(snake_case_,json={"""text""": message_body},headers=snake_case_ )
if response.status_code != 200:
_A : Union[str, Any] = (
"""Request to slack returned an error """
f'''{response.status_code}, the response is:\n{response.text}'''
)
raise ValueError(snake_case_ )
if __name__ == "__main__":
# Set the slack url to the one provided by Slack when you create the webhook at
# https://my.slack.com/services/new/incoming-webhook/
send_slack_message("<YOUR MESSAGE BODY>", "<SLACK CHANNEL URL>")
| 26 |
def lowerCAmelCase_ ( snake_case_ ):
if number < 0:
raise ValueError("""number must not be negative""" )
return number & (number - 1) == 0
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available
_snake_case = {
"configuration_nllb_moe": [
"NLLB_MOE_PRETRAINED_CONFIG_ARCHIVE_MAP",
"NllbMoeConfig",
]
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"NLLB_MOE_PRETRAINED_MODEL_ARCHIVE_LIST",
"NllbMoeForConditionalGeneration",
"NllbMoeModel",
"NllbMoePreTrainedModel",
"NllbMoeTop2Router",
"NllbMoeSparseMLP",
]
if TYPE_CHECKING:
from .configuration_nllb_moe import (
NLLB_MOE_PRETRAINED_CONFIG_ARCHIVE_MAP,
NllbMoeConfig,
)
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_nllb_moe import (
NLLB_MOE_PRETRAINED_MODEL_ARCHIVE_LIST,
NllbMoeForConditionalGeneration,
NllbMoeModel,
NllbMoePreTrainedModel,
NllbMoeSparseMLP,
NllbMoeTopaRouter,
)
else:
import sys
_snake_case = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
| 26 |
import argparse
import torch
from transformers import BlenderbotConfig, BlenderbotForConditionalGeneration
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
_snake_case = [
["attention", "attn"],
["encoder_attention", "encoder_attn"],
["q_lin", "q_proj"],
["k_lin", "k_proj"],
["v_lin", "v_proj"],
["out_lin", "out_proj"],
["norm_embeddings", "layernorm_embedding"],
["position_embeddings", "embed_positions"],
["embeddings", "embed_tokens"],
["ffn.lin", "fc"],
]
def lowerCAmelCase_ ( snake_case_ ):
if k == "embeddings.weight":
return "shared.weight"
for parlai_name, hf_name in PATTERNS:
_A : str = k.replace(snake_case_,snake_case_ )
if k.startswith("""encoder""" ):
_A : Optional[Any] = k.replace(""".attn""",""".self_attn""" )
_A : Dict = k.replace("""norm1""","""self_attn_layer_norm""" )
_A : Optional[Any] = k.replace("""norm2""","""final_layer_norm""" )
elif k.startswith("""decoder""" ):
_A : str = k.replace("""norm1""","""self_attn_layer_norm""" )
_A : Any = k.replace("""norm2""","""encoder_attn_layer_norm""" )
_A : Optional[int] = k.replace("""norm3""","""final_layer_norm""" )
return k
def lowerCAmelCase_ ( snake_case_ ):
_A : List[Any] = [
"""model.encoder.layernorm_embedding.weight""",
"""model.encoder.layernorm_embedding.bias""",
"""model.decoder.layernorm_embedding.weight""",
"""model.decoder.layernorm_embedding.bias""",
]
for k in keys:
_A : str = sd.pop(snake_case_ )
_A : Optional[int] = k.replace("""layernorm_embedding""","""layer_norm""" )
assert new_k not in sd
_A : Optional[int] = v
_snake_case = ["START"]
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
_A : List[Any] = model["""model"""]
_A : Optional[Any] = BlenderbotConfig.from_json_file(snake_case_ )
_A : List[str] = BlenderbotForConditionalGeneration(snake_case_ )
_A : Tuple = m.model.state_dict().keys()
_A : Any = []
_A : Dict = {}
for k, v in sd.items():
if k in IGNORE_KEYS:
continue
_A : Optional[int] = rename_state_dict_key(snake_case_ )
if new_k not in valid_keys:
failures.append([k, new_k] )
else:
_A : Dict = v
if cfg.normalize_before: # Blenderbot-3B checkpoints. Rename layernorm_embedding -> layer_norm
rename_layernorm_keys(snake_case_ )
m.model.load_state_dict(snake_case_,strict=snake_case_ )
m.half()
m.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument("--src_path", type=str, help="like blenderbot-model.bin")
parser.add_argument("--save_dir", default="hf_blenderbot", type=str, help="Where to save converted model.")
parser.add_argument(
"--hf_config_json", default="blenderbot-3b-config.json", type=str, help="Path to config to use"
)
_snake_case = parser.parse_args()
convert_parlai_checkpoint(args.src_path, args.save_dir, args.hf_config_json)
| 26 | 1 |
import logging
import math
from functools import partial
from typing import Any, Callable, Dict, Iterable, List, Optional, Sequence, Tuple, Union
import torch
from .tensor_utils import tensor_tree_map, tree_map
def lowerCAmelCase_ ( snake_case_ ):
_A : Optional[int] = []
if isinstance(snake_case_,snake_case_ ):
for v in tree.values():
shapes.extend(_fetch_dims(snake_case_ ) )
elif isinstance(snake_case_,(list, tuple) ):
for t in tree:
shapes.extend(_fetch_dims(snake_case_ ) )
elif isinstance(snake_case_,torch.Tensor ):
shapes.append(tree.shape )
else:
raise ValueError("""Not supported""" )
return shapes
@torch.jit.ignore
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : int = []
for d in reversed(snake_case_ ):
idx.append(flat_idx % d )
_A : List[str] = flat_idx // d
return tuple(reversed(snake_case_ ) )
@torch.jit.ignore
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ = None,snake_case_ = None,):
# start_edges and end_edges both indicate whether, starting from any given
# dimension, the start/end index is at the top/bottom edge of the
# corresponding tensor, modeled as a tree
def reduce_edge_list(snake_case_ ) -> None:
_A : List[str] = True
for i in range(len(snake_case_ ) ):
_A : List[str] = -1 * (i + 1)
l[reversed_idx] &= tally
_A : List[Any] = l[reversed_idx]
if start_edges is None:
_A : str = [s == 0 for s in start]
reduce_edge_list(snake_case_ )
if end_edges is None:
_A : Any = [e == (d - 1) for e, d in zip(snake_case_,snake_case_ )]
reduce_edge_list(snake_case_ )
# Base cases. Either start/end are empty and we're done, or the final,
# one-dimensional tensor can be simply sliced
if len(snake_case_ ) == 0:
return [()]
elif len(snake_case_ ) == 1:
return [(slice(start[0],end[0] + 1 ),)]
_A : List[Tuple[slice, ...]] = []
_A : List[slice] = []
# Dimensions common to start and end can be selected directly
for s, e in zip(snake_case_,snake_case_ ):
if s == e:
path_list.append(slice(snake_case_,s + 1 ) )
else:
break
_A : Tuple[slice, ...] = tuple(snake_case_ )
_A : str = len(snake_case_ )
# start == end, and we're done
if divergence_idx == len(snake_case_ ):
return [path]
def upper() -> Tuple[Tuple[slice, ...], ...]:
assert start_edges is not None
assert end_edges is not None
_A : Any = start[divergence_idx]
return tuple(
path + (slice(snake_case_,sdi + 1 ),) + s
for s in _get_minimal_slice_set(
start[divergence_idx + 1 :],[d - 1 for d in dims[divergence_idx + 1 :]],dims[divergence_idx + 1 :],start_edges=start_edges[divergence_idx + 1 :],end_edges=[True for _ in end_edges[divergence_idx + 1 :]],) )
def lower() -> Tuple[Tuple[slice, ...], ...]:
assert start_edges is not None
assert end_edges is not None
_A : Union[str, Any] = end[divergence_idx]
return tuple(
path + (slice(snake_case_,edi + 1 ),) + s
for s in _get_minimal_slice_set(
[0 for _ in start[divergence_idx + 1 :]],end[divergence_idx + 1 :],dims[divergence_idx + 1 :],start_edges=[True for _ in start_edges[divergence_idx + 1 :]],end_edges=end_edges[divergence_idx + 1 :],) )
# If both start and end are at the edges of the subtree rooted at
# divergence_idx, we can just select the whole subtree at once
if start_edges[divergence_idx] and end_edges[divergence_idx]:
slices.append(path + (slice(start[divergence_idx],end[divergence_idx] + 1 ),) )
# If just start is at the edge, we can grab almost all of the subtree,
# treating only the ragged bottom edge as an edge case
elif start_edges[divergence_idx]:
slices.append(path + (slice(start[divergence_idx],end[divergence_idx] ),) )
slices.extend(lower() )
# Analogous to the previous case, but the top is ragged this time
elif end_edges[divergence_idx]:
slices.extend(upper() )
slices.append(path + (slice(start[divergence_idx] + 1,end[divergence_idx] + 1 ),) )
# If both sides of the range are ragged, we need to handle both sides
# separately. If there's contiguous meat in between them, we can index it
# in one big chunk
else:
slices.extend(upper() )
_A : List[str] = end[divergence_idx] - start[divergence_idx]
if middle_ground > 1:
slices.append(path + (slice(start[divergence_idx] + 1,end[divergence_idx] ),) )
slices.extend(lower() )
return slices
@torch.jit.ignore
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : Optional[int] = t.shape[:no_batch_dims]
_A : Tuple = list(_flat_idx_to_idx(snake_case_,snake_case_ ) )
# _get_minimal_slice_set is inclusive
_A : str = list(_flat_idx_to_idx(flat_end - 1,snake_case_ ) )
# Get an ordered list of slices to perform
_A : int = _get_minimal_slice_set(
snake_case_,snake_case_,snake_case_,)
_A : Optional[int] = [t[s] for s in slices]
return torch.cat([s.view((-1,) + t.shape[no_batch_dims:] ) for s in sliced_tensors] )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_,snake_case_ = False,snake_case_ = None,snake_case_ = False,):
if not (len(snake_case_ ) > 0):
raise ValueError("""Must provide at least one input""" )
_A : int = [shape[:no_batch_dims] for shape in _fetch_dims(snake_case_ )]
_A : Tuple = tuple([max(snake_case_ ) for s in zip(*snake_case_ )] )
def _prep_inputs(snake_case_ ) -> torch.Tensor:
if not low_mem:
if not sum(t.shape[:no_batch_dims] ) == no_batch_dims:
_A : Dict = t.expand(orig_batch_dims + t.shape[no_batch_dims:] )
_A : Union[str, Any] = t.reshape(-1,*t.shape[no_batch_dims:] )
else:
_A : Dict = t.expand(orig_batch_dims + t.shape[no_batch_dims:] )
return t
_A : Dict[str, Any] = tensor_tree_map(_prep_inputs,snake_case_ )
_A : Tuple = None
if _out is not None:
_A : Optional[int] = tensor_tree_map(lambda snake_case_ : t.view([-1] + list(t.shape[no_batch_dims:] ) ),_out )
_A : Optional[Any] = 1
for d in orig_batch_dims:
flat_batch_dim *= d
_A : str = flat_batch_dim // chunk_size + (flat_batch_dim % chunk_size != 0)
def _select_chunk(snake_case_ ) -> torch.Tensor:
return t[i : i + chunk_size] if t.shape[0] != 1 else t
_A : Any = 0
_A : List[str] = prepped_outputs
for _ in range(snake_case_ ):
# Chunk the input
if not low_mem:
_A : List[Any] = _select_chunk
else:
_A : str = partial(
_chunk_slice,flat_start=snake_case_,flat_end=min(snake_case_,i + chunk_size ),no_batch_dims=len(snake_case_ ),)
_A : Dict[str, Any] = tensor_tree_map(snake_case_,snake_case_ )
# Run the layer on the chunk
_A : Tuple = layer(**snake_case_ )
# Allocate space for the output
if out is None:
_A : Optional[Any] = tensor_tree_map(lambda snake_case_ : t.new_zeros((flat_batch_dim,) + t.shape[1:] ),snake_case_ )
# Put the chunk in its pre-allocated space
if isinstance(snake_case_,snake_case_ ):
def assign(snake_case_,snake_case_ ) -> None:
for k, v in da.items():
if isinstance(snake_case_,snake_case_ ):
assign(snake_case_,da[k] )
else:
if _add_into_out:
v[i : i + chunk_size] += da[k]
else:
_A : List[Any] = da[k]
assign(snake_case_,snake_case_ )
elif isinstance(snake_case_,snake_case_ ):
for xa, xa in zip(snake_case_,snake_case_ ):
if _add_into_out:
xa[i : i + chunk_size] += xa
else:
_A : Tuple = xa
elif isinstance(snake_case_,torch.Tensor ):
if _add_into_out:
out[i : i + chunk_size] += output_chunk
else:
_A : Optional[int] = output_chunk
else:
raise ValueError("""Not supported""" )
i += chunk_size
_A : int = tensor_tree_map(lambda snake_case_ : t.view(orig_batch_dims + t.shape[1:] ),snake_case_ )
return out
class lowercase :
def __init__( self , _a = 512 , ) -> str:
_A : Union[str, Any] = max_chunk_size
_A : Optional[int] = None
_A : Optional[tuple] = None
def a__ ( self , _a , _a , _a ) -> int:
logging.info("""Tuning chunk size...""" )
if min_chunk_size >= self.max_chunk_size:
return min_chunk_size
_A : List[int] = [2**l for l in range(int(math.log(self.max_chunk_size , 2 ) ) + 1 )]
_A : List[Any] = [c for c in candidates if c > min_chunk_size]
_A : Optional[Any] = [min_chunk_size] + candidates
candidates[-1] += 4
def test_chunk_size(_a ) -> bool:
try:
with torch.no_grad():
fn(*_a , chunk_size=_a )
return True
except RuntimeError:
return False
_A : Optional[int] = 0
_A : str = len(_a ) - 1
while i > min_viable_chunk_size_index:
_A : Optional[Any] = test_chunk_size(candidates[i] )
if not viable:
_A : Union[str, Any] = (min_viable_chunk_size_index + i) // 2
else:
_A : Union[str, Any] = i
_A : Optional[Any] = (i + len(_a ) - 1) // 2
return candidates[min_viable_chunk_size_index]
def a__ ( self , _a , _a ) -> bool:
_A : Optional[Any] = True
for aa, aa in zip(_a , _a ):
assert type(_a ) == type(_a )
if isinstance(_a , (list, tuple) ):
consistent &= self._compare_arg_caches(_a , _a )
elif isinstance(_a , _a ):
_A : Union[str, Any] = [v for _, v in sorted(aa.items() , key=lambda _a : x[0] )]
_A : Union[str, Any] = [v for _, v in sorted(aa.items() , key=lambda _a : x[0] )]
consistent &= self._compare_arg_caches(_a , _a )
else:
consistent &= aa == aa
return consistent
def a__ ( self , _a , _a , _a , ) -> int:
_A : Tuple = True
_A : tuple = tree_map(lambda _a : a.shape if isinstance(_a , torch.Tensor ) else a , _a , _a )
if self.cached_arg_data is not None:
# If args have changed shape/value, we need to re-tune
assert len(self.cached_arg_data ) == len(_a )
_A : List[Any] = self._compare_arg_caches(self.cached_arg_data , _a )
else:
# Otherwise, we can reuse the precomputed value
_A : Tuple = False
if not consistent:
_A : int = self._determine_favorable_chunk_size(
_a , _a , _a , )
_A : int = arg_data
assert self.cached_chunk_size is not None
return self.cached_chunk_size
| 26 |
import multiprocessing
from typing import TYPE_CHECKING, Optional, Union
from .. import Dataset, Features, config
from ..formatting import query_table
from ..packaged_modules.sql.sql import Sql
from ..utils import logging
from .abc import AbstractDatasetInputStream
if TYPE_CHECKING:
import sqlitea
import sqlalchemy
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a = None , _a = None , _a = False , **_a , ) -> int:
super().__init__(features=_a , cache_dir=_a , keep_in_memory=_a , **_a )
_A : Optional[int] = Sql(
cache_dir=_a , features=_a , sql=_a , con=_a , **_a , )
def a__ ( self ) -> Optional[Any]:
_A : Tuple = None
_A : int = None
_A : Tuple = None
_A : Union[str, Any] = None
self.builder.download_and_prepare(
download_config=_a , download_mode=_a , verification_mode=_a , base_path=_a , )
# Build dataset for splits
_A : int = self.builder.as_dataset(
split="""train""" , verification_mode=_a , in_memory=self.keep_in_memory )
return dataset
class lowercase :
def __init__( self , _a , _a , _a , _a = None , _a = None , **_a , ) -> Union[str, Any]:
if num_proc is not None and num_proc <= 0:
raise ValueError(F'''num_proc {num_proc} must be an integer > 0.''' )
_A : Dict = dataset
_A : int = name
_A : Union[str, Any] = con
_A : str = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE
_A : str = num_proc
_A : Optional[Any] = to_sql_kwargs
def a__ ( self ) -> int:
_A : Any = self.to_sql_kwargs.pop("""sql""" , _a )
_A : List[str] = self.to_sql_kwargs.pop("""con""" , _a )
_A : int = self.to_sql_kwargs.pop("""index""" , _a )
_A : List[str] = self._write(index=_a , **self.to_sql_kwargs )
return written
def a__ ( self , _a ) -> Optional[int]:
_A , _A , _A : List[str] = args
_A : int = {**to_sql_kwargs, """if_exists""": """append"""} if offset > 0 else to_sql_kwargs
_A : str = query_table(
table=self.dataset.data , key=slice(_a , offset + self.batch_size ) , indices=self.dataset._indices , )
_A : Tuple = batch.to_pandas()
_A : Union[str, Any] = df.to_sql(self.name , self.con , index=_a , **_a )
return num_rows or len(_a )
def a__ ( self , _a , **_a ) -> int:
_A : Any = 0
if self.num_proc is None or self.num_proc == 1:
for offset in logging.tqdm(
range(0 , len(self.dataset ) , self.batch_size ) , unit="""ba""" , disable=not logging.is_progress_bar_enabled() , desc="""Creating SQL from Arrow format""" , ):
written += self._batch_sql((offset, index, to_sql_kwargs) )
else:
_A , _A : Tuple = len(self.dataset ), self.batch_size
with multiprocessing.Pool(self.num_proc ) as pool:
for num_rows in logging.tqdm(
pool.imap(
self._batch_sql , [(offset, index, to_sql_kwargs) for offset in range(0 , _a , _a )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit="""ba""" , disable=not logging.is_progress_bar_enabled() , desc="""Creating SQL from Arrow format""" , ):
written += num_rows
return written
| 26 | 1 |
import numpy as np
import torch
import tqdm
from ...models.unet_ad import UNetaDModel
from ...pipelines import DiffusionPipeline
from ...utils import randn_tensor
from ...utils.dummy_pt_objects import DDPMScheduler
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a , _a , ) -> int:
super().__init__()
_A : List[Any] = value_function
_A : Union[str, Any] = unet
_A : str = scheduler
_A : Union[str, Any] = env
_A : Tuple = env.get_dataset()
_A : Optional[int] = {}
for key in self.data.keys():
try:
_A : Any = self.data[key].mean()
except: # noqa: E722
pass
_A : Union[str, Any] = {}
for key in self.data.keys():
try:
_A : Optional[Any] = self.data[key].std()
except: # noqa: E722
pass
_A : str = env.observation_space.shape[0]
_A : Union[str, Any] = env.action_space.shape[0]
def a__ ( self , _a , _a ) -> Union[str, Any]:
return (x_in - self.means[key]) / self.stds[key]
def a__ ( self , _a , _a ) -> Optional[int]:
return x_in * self.stds[key] + self.means[key]
def a__ ( self , _a ) -> List[str]:
if type(_a ) is dict:
return {k: self.to_torch(_a ) for k, v in x_in.items()}
elif torch.is_tensor(_a ):
return x_in.to(self.unet.device )
return torch.tensor(_a , device=self.unet.device )
def a__ ( self , _a , _a , _a ) -> Optional[Any]:
for key, val in cond.items():
_A : int = val.clone()
return x_in
def a__ ( self , _a , _a , _a , _a ) -> Optional[int]:
_A : Union[str, Any] = x.shape[0]
_A : Any = None
for i in tqdm.tqdm(self.scheduler.timesteps ):
# create batch of timesteps to pass into model
_A : Dict = torch.full((batch_size,) , _a , device=self.unet.device , dtype=torch.long )
for _ in range(_a ):
with torch.enable_grad():
x.requires_grad_()
# permute to match dimension for pre-trained models
_A : List[Any] = self.value_function(x.permute(0 , 2 , 1 ) , _a ).sample
_A : Optional[Any] = torch.autograd.grad([y.sum()] , [x] )[0]
_A : Optional[int] = self.scheduler._get_variance(_a )
_A : List[Any] = torch.exp(0.5 * posterior_variance )
_A : Tuple = model_std * grad
_A : Optional[Any] = 0
_A : Any = x.detach()
_A : List[Any] = x + scale * grad
_A : Any = self.reset_xa(_a , _a , self.action_dim )
_A : str = self.unet(x.permute(0 , 2 , 1 ) , _a ).sample.permute(0 , 2 , 1 )
# TODO: verify deprecation of this kwarg
_A : Dict = self.scheduler.step(_a , _a , _a , predict_epsilon=_a )["""prev_sample"""]
# apply conditions to the trajectory (set the initial state)
_A : str = self.reset_xa(_a , _a , self.action_dim )
_A : List[Any] = self.to_torch(_a )
return x, y
def __call__( self , _a , _a=64 , _a=32 , _a=2 , _a=0.1 ) -> Any:
# normalize the observations and create batch dimension
_A : Any = self.normalize(_a , """observations""" )
_A : Optional[int] = obs[None].repeat(_a , axis=0 )
_A : List[str] = {0: self.to_torch(_a )}
_A : int = (batch_size, planning_horizon, self.state_dim + self.action_dim)
# generate initial noise and apply our conditions (to make the trajectories start at current state)
_A : Optional[int] = randn_tensor(_a , device=self.unet.device )
_A : int = self.reset_xa(_a , _a , self.action_dim )
_A : int = self.to_torch(_a )
# run the diffusion process
_A , _A : List[Any] = self.run_diffusion(_a , _a , _a , _a )
# sort output trajectories by value
_A : Optional[int] = y.argsort(0 , descending=_a ).squeeze()
_A : Tuple = x[sorted_idx]
_A : Optional[int] = sorted_values[:, :, : self.action_dim]
_A : Union[str, Any] = actions.detach().cpu().numpy()
_A : Tuple = self.de_normalize(_a , key="""actions""" )
# select the action with the highest value
if y is not None:
_A : int = 0
else:
# if we didn't run value guiding, select a random action
_A : Optional[Any] = np.random.randint(0 , _a )
_A : Any = denorm_actions[selected_index, 0]
return denorm_actions
| 26 |
from ...configuration_utils import PretrainedConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/config.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/config.json"
# See all FNet models at https://huggingface.co/models?filter=fnet
}
class lowercase ( UpperCamelCase__ ):
_a = "fnet"
def __init__( self , _a=3_2000 , _a=768 , _a=12 , _a=3072 , _a="gelu_new" , _a=0.1 , _a=512 , _a=4 , _a=0.02 , _a=1e-12 , _a=False , _a=512 , _a=3 , _a=1 , _a=2 , **_a , ) -> int:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Any = vocab_size
_A : str = max_position_embeddings
_A : Optional[Any] = hidden_size
_A : List[str] = num_hidden_layers
_A : List[str] = intermediate_size
_A : List[Any] = hidden_act
_A : List[str] = hidden_dropout_prob
_A : List[str] = initializer_range
_A : List[Any] = type_vocab_size
_A : List[Any] = layer_norm_eps
_A : List[str] = use_tpu_fourier_optimizations
_A : str = tpu_short_seq_length
| 26 | 1 |
import pyarrow.parquet as pq
import pytest
from datasets import Audio, Dataset, DatasetDict, Features, NamedSplit, Sequence, Value, config
from datasets.features.image import Image
from datasets.io.parquet import ParquetDatasetReader, ParquetDatasetWriter, get_writer_batch_size
from ..utils import assert_arrow_memory_doesnt_increase, assert_arrow_memory_increases
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
assert isinstance(snake_case_,snake_case_ )
assert dataset.num_rows == 4
assert dataset.num_columns == 3
assert dataset.column_names == ["col_1", "col_2", "col_3"]
for feature, expected_dtype in expected_features.items():
assert dataset.features[feature].dtype == expected_dtype
@pytest.mark.parametrize("""keep_in_memory""",[False, True] )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Optional[int] = tmp_path / """cache"""
_A : Optional[int] = {"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""}
with assert_arrow_memory_increases() if keep_in_memory else assert_arrow_memory_doesnt_increase():
_A : Dict = ParquetDatasetReader(snake_case_,cache_dir=snake_case_,keep_in_memory=snake_case_ ).read()
_check_parquet_dataset(snake_case_,snake_case_ )
@pytest.mark.parametrize(
"""features""",[
None,
{"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""},
{"""col_1""": """string""", """col_2""": """string""", """col_3""": """string"""},
{"""col_1""": """int32""", """col_2""": """int32""", """col_3""": """int32"""},
{"""col_1""": """float32""", """col_2""": """float32""", """col_3""": """float32"""},
],)
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Union[str, Any] = tmp_path / """cache"""
_A : Optional[Any] = {"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""}
_A : Optional[int] = features.copy() if features else default_expected_features
_A : Tuple = (
Features({feature: Value(snake_case_ ) for feature, dtype in features.items()} ) if features is not None else None
)
_A : Optional[int] = ParquetDatasetReader(snake_case_,features=snake_case_,cache_dir=snake_case_ ).read()
_check_parquet_dataset(snake_case_,snake_case_ )
@pytest.mark.parametrize("""split""",[None, NamedSplit("""train""" ), """train""", """test"""] )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Dict = tmp_path / """cache"""
_A : int = {"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""}
_A : Optional[int] = ParquetDatasetReader(snake_case_,cache_dir=snake_case_,split=snake_case_ ).read()
_check_parquet_dataset(snake_case_,snake_case_ )
assert dataset.split == split if split else "train"
@pytest.mark.parametrize("""path_type""",[str, list] )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
if issubclass(snake_case_,snake_case_ ):
_A : str = parquet_path
elif issubclass(snake_case_,snake_case_ ):
_A : List[Any] = [parquet_path]
_A : Tuple = tmp_path / """cache"""
_A : List[str] = {"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""}
_A : str = ParquetDatasetReader(snake_case_,cache_dir=snake_case_ ).read()
_check_parquet_dataset(snake_case_,snake_case_ )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_=("train",) ):
assert isinstance(snake_case_,snake_case_ )
for split in splits:
_A : str = dataset_dict[split]
assert dataset.num_rows == 4
assert dataset.num_columns == 3
assert dataset.column_names == ["col_1", "col_2", "col_3"]
for feature, expected_dtype in expected_features.items():
assert dataset.features[feature].dtype == expected_dtype
@pytest.mark.parametrize("""keep_in_memory""",[False, True] )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Optional[Any] = tmp_path / """cache"""
_A : Any = {"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""}
with assert_arrow_memory_increases() if keep_in_memory else assert_arrow_memory_doesnt_increase():
_A : List[str] = ParquetDatasetReader(
{"""train""": parquet_path},cache_dir=snake_case_,keep_in_memory=snake_case_ ).read()
_check_parquet_datasetdict(snake_case_,snake_case_ )
@pytest.mark.parametrize(
"""features""",[
None,
{"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""},
{"""col_1""": """string""", """col_2""": """string""", """col_3""": """string"""},
{"""col_1""": """int32""", """col_2""": """int32""", """col_3""": """int32"""},
{"""col_1""": """float32""", """col_2""": """float32""", """col_3""": """float32"""},
],)
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Union[str, Any] = tmp_path / """cache"""
_A : Any = {"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""}
_A : str = features.copy() if features else default_expected_features
_A : List[Any] = (
Features({feature: Value(snake_case_ ) for feature, dtype in features.items()} ) if features is not None else None
)
_A : Optional[int] = ParquetDatasetReader({"""train""": parquet_path},features=snake_case_,cache_dir=snake_case_ ).read()
_check_parquet_datasetdict(snake_case_,snake_case_ )
@pytest.mark.parametrize("""split""",[None, NamedSplit("""train""" ), """train""", """test"""] )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
if split:
_A : List[str] = {split: parquet_path}
else:
_A : List[Any] = """train"""
_A : Union[str, Any] = {"""train""": parquet_path, """test""": parquet_path}
_A : str = tmp_path / """cache"""
_A : Dict = {"""col_1""": """string""", """col_2""": """int64""", """col_3""": """float64"""}
_A : Union[str, Any] = ParquetDatasetReader(snake_case_,cache_dir=snake_case_ ).read()
_check_parquet_datasetdict(snake_case_,snake_case_,splits=list(path.keys() ) )
assert all(dataset[split].split == split for split in path.keys() )
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Dict = ParquetDatasetWriter(snake_case_,tmp_path / """foo.parquet""" )
assert writer.write() > 0
_A : List[str] = pq.ParquetFile(tmp_path / """foo.parquet""" )
_A : Dict = pf.read()
assert dataset.data.table == output_table
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : int = str(shared_datadir / """test_image_rgb.jpg""" )
_A : Any = {"""image""": [image_path]}
_A : Union[str, Any] = Features({"""image""": Image()} )
_A : Union[str, Any] = Dataset.from_dict(snake_case_,features=snake_case_ )
_A : str = ParquetDatasetWriter(snake_case_,tmp_path / """foo.parquet""" )
assert writer.write() > 0
_A : int = Dataset.from_parquet(str(tmp_path / """foo.parquet""" ) )
assert dataset.features == reloaded_dataset.features
_A : int = ParquetDatasetReader(str(tmp_path / """foo.parquet""" ),streaming=snake_case_ ).read()
assert dataset.features == reloaded_iterable_dataset.features
@pytest.mark.parametrize(
"""feature, expected""",[
(Features({"""foo""": Value("""int32""" )} ), None),
(Features({"""image""": Image(), """foo""": Value("""int32""" )} ), config.PARQUET_ROW_GROUP_SIZE_FOR_IMAGE_DATASETS),
(Features({"""nested""": Sequence(Audio() )} ), config.PARQUET_ROW_GROUP_SIZE_FOR_AUDIO_DATASETS),
],)
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
assert get_writer_batch_size(snake_case_ ) == expected
| 26 |
def lowerCAmelCase_ ( snake_case_ ):
if n_term == "":
return []
_A : list = []
for temp in range(int(snake_case_ ) ):
series.append(f'''1/{temp + 1}''' if series else """1""" )
return series
if __name__ == "__main__":
_snake_case = input("Enter the last number (nth term) of the Harmonic Series")
print("Formula of Harmonic Series => 1+1/2+1/3 ..... 1/n")
print(harmonic_series(nth_term))
| 26 | 1 |
import argparse
import json
import os
import pickle
import shutil
import numpy as np
import torch
from distiller import Distiller
from lm_seqs_dataset import LmSeqsDataset
from transformers import (
BertConfig,
BertForMaskedLM,
BertTokenizer,
DistilBertConfig,
DistilBertForMaskedLM,
DistilBertTokenizer,
GPTaConfig,
GPTaLMHeadModel,
GPTaTokenizer,
RobertaConfig,
RobertaForMaskedLM,
RobertaTokenizer,
)
from utils import git_log, init_gpu_params, logger, set_seed
_snake_case = {
"distilbert": (DistilBertConfig, DistilBertForMaskedLM, DistilBertTokenizer),
"roberta": (RobertaConfig, RobertaForMaskedLM, RobertaTokenizer),
"bert": (BertConfig, BertForMaskedLM, BertTokenizer),
"gpt2": (GPTaConfig, GPTaLMHeadModel, GPTaTokenizer),
}
def lowerCAmelCase_ ( snake_case_ ):
assert (args.mlm and args.alpha_mlm > 0.0) or (not args.mlm and args.alpha_mlm == 0.0)
assert (args.alpha_mlm > 0.0 and args.alpha_clm == 0.0) or (args.alpha_mlm == 0.0 and args.alpha_clm > 0.0)
if args.mlm:
assert os.path.isfile(args.token_counts )
assert (args.student_type in ["roberta", "distilbert"]) and (args.teacher_type in ["roberta", "bert"])
else:
assert (args.student_type in ["gpt2"]) and (args.teacher_type in ["gpt2"])
assert args.teacher_type == args.student_type or (
args.student_type == "distilbert" and args.teacher_type == "bert"
)
assert os.path.isfile(args.student_config )
if args.student_pretrained_weights is not None:
assert os.path.isfile(args.student_pretrained_weights )
if args.freeze_token_type_embds:
assert args.student_type in ["roberta"]
assert args.alpha_ce >= 0.0
assert args.alpha_mlm >= 0.0
assert args.alpha_clm >= 0.0
assert args.alpha_mse >= 0.0
assert args.alpha_cos >= 0.0
assert args.alpha_ce + args.alpha_mlm + args.alpha_clm + args.alpha_mse + args.alpha_cos > 0.0
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if args.student_type == "roberta":
_A : List[str] = False
elif args.student_type == "gpt2":
_A : Optional[int] = False
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if args.student_type == "roberta":
_A : str = False
def lowerCAmelCase_ ( ):
_A : Union[str, Any] = argparse.ArgumentParser(description="""Training""" )
parser.add_argument("""--force""",action="""store_true""",help="""Overwrite dump_path if it already exists.""" )
parser.add_argument(
"""--dump_path""",type=snake_case_,required=snake_case_,help="""The output directory (log, checkpoints, parameters, etc.)""" )
parser.add_argument(
"""--data_file""",type=snake_case_,required=snake_case_,help="""The binarized file (tokenized + tokens_to_ids) and grouped by sequence.""",)
parser.add_argument(
"""--student_type""",type=snake_case_,choices=["""distilbert""", """roberta""", """gpt2"""],required=snake_case_,help="""The student type (DistilBERT, RoBERTa).""",)
parser.add_argument("""--student_config""",type=snake_case_,required=snake_case_,help="""Path to the student configuration.""" )
parser.add_argument(
"""--student_pretrained_weights""",default=snake_case_,type=snake_case_,help="""Load student initialization checkpoint.""" )
parser.add_argument(
"""--teacher_type""",choices=["""bert""", """roberta""", """gpt2"""],required=snake_case_,help="""Teacher type (BERT, RoBERTa).""" )
parser.add_argument("""--teacher_name""",type=snake_case_,required=snake_case_,help="""The teacher model.""" )
parser.add_argument("""--temperature""",default=2.0,type=snake_case_,help="""Temperature for the softmax temperature.""" )
parser.add_argument(
"""--alpha_ce""",default=0.5,type=snake_case_,help="""Linear weight for the distillation loss. Must be >=0.""" )
parser.add_argument(
"""--alpha_mlm""",default=0.0,type=snake_case_,help="""Linear weight for the MLM loss. Must be >=0. Should be used in conjunction with `mlm` flag.""",)
parser.add_argument("""--alpha_clm""",default=0.5,type=snake_case_,help="""Linear weight for the CLM loss. Must be >=0.""" )
parser.add_argument("""--alpha_mse""",default=0.0,type=snake_case_,help="""Linear weight of the MSE loss. Must be >=0.""" )
parser.add_argument(
"""--alpha_cos""",default=0.0,type=snake_case_,help="""Linear weight of the cosine embedding loss. Must be >=0.""" )
parser.add_argument(
"""--mlm""",action="""store_true""",help="""The LM step: MLM or CLM. If `mlm` is True, the MLM is used over CLM.""" )
parser.add_argument(
"""--mlm_mask_prop""",default=0.15,type=snake_case_,help="""Proportion of tokens for which we need to make a prediction.""",)
parser.add_argument("""--word_mask""",default=0.8,type=snake_case_,help="""Proportion of tokens to mask out.""" )
parser.add_argument("""--word_keep""",default=0.1,type=snake_case_,help="""Proportion of tokens to keep.""" )
parser.add_argument("""--word_rand""",default=0.1,type=snake_case_,help="""Proportion of tokens to randomly replace.""" )
parser.add_argument(
"""--mlm_smoothing""",default=0.7,type=snake_case_,help="""Smoothing parameter to emphasize more rare tokens (see XLM, similar to word2vec).""",)
parser.add_argument("""--token_counts""",type=snake_case_,help="""The token counts in the data_file for MLM.""" )
parser.add_argument(
"""--restrict_ce_to_mask""",action="""store_true""",help="""If true, compute the distillation loss only the [MLM] prediction distribution.""",)
parser.add_argument(
"""--freeze_pos_embs""",action="""store_true""",help="""Freeze positional embeddings during distillation. For student_type in ['roberta', 'gpt2'] only.""",)
parser.add_argument(
"""--freeze_token_type_embds""",action="""store_true""",help="""Freeze token type embeddings during distillation if existent. For student_type in ['roberta'] only.""",)
parser.add_argument("""--n_epoch""",type=snake_case_,default=3,help="""Number of pass on the whole dataset.""" )
parser.add_argument("""--batch_size""",type=snake_case_,default=5,help="""Batch size (for each process).""" )
parser.add_argument(
"""--group_by_size""",action="""store_false""",help="""If true, group sequences that have similar length into the same batch. Default is true.""",)
parser.add_argument(
"""--gradient_accumulation_steps""",type=snake_case_,default=50,help="""Gradient accumulation for larger training batches.""",)
parser.add_argument("""--warmup_prop""",default=0.05,type=snake_case_,help="""Linear warmup proportion.""" )
parser.add_argument("""--weight_decay""",default=0.0,type=snake_case_,help="""Weight decay if we apply some.""" )
parser.add_argument("""--learning_rate""",default=5e-4,type=snake_case_,help="""The initial learning rate for Adam.""" )
parser.add_argument("""--adam_epsilon""",default=1e-6,type=snake_case_,help="""Epsilon for Adam optimizer.""" )
parser.add_argument("""--max_grad_norm""",default=5.0,type=snake_case_,help="""Max gradient norm.""" )
parser.add_argument("""--initializer_range""",default=0.02,type=snake_case_,help="""Random initialization range.""" )
parser.add_argument(
"""--fp16""",action="""store_true""",help="""Whether to use 16-bit (mixed) precision (through NVIDIA apex) instead of 32-bit""",)
parser.add_argument(
"""--fp16_opt_level""",type=snake_case_,default="""O1""",help=(
"""For fp16: Apex AMP optimization level selected in ['O0', 'O1', 'O2', and 'O3']."""
"""See details at https://nvidia.github.io/apex/amp.html"""
),)
parser.add_argument("""--n_gpu""",type=snake_case_,default=1,help="""Number of GPUs in the node.""" )
parser.add_argument("""--local_rank""",type=snake_case_,default=-1,help="""Distributed training - Local rank""" )
parser.add_argument("""--seed""",type=snake_case_,default=56,help="""Random seed""" )
parser.add_argument("""--log_interval""",type=snake_case_,default=500,help="""Tensorboard logging interval.""" )
parser.add_argument("""--checkpoint_interval""",type=snake_case_,default=4000,help="""Checkpoint interval.""" )
_A : List[str] = parser.parse_args()
sanity_checks(snake_case_ )
# ARGS #
init_gpu_params(snake_case_ )
set_seed(snake_case_ )
if args.is_master:
if os.path.exists(args.dump_path ):
if not args.force:
raise ValueError(
f'''Serialization dir {args.dump_path} already exists, but you have not precised wheter to overwrite'''
""" itUse `--force` if you want to overwrite it""" )
else:
shutil.rmtree(args.dump_path )
if not os.path.exists(args.dump_path ):
os.makedirs(args.dump_path )
logger.info(f'''Experiment will be dumped and logged in {args.dump_path}''' )
# SAVE PARAMS #
logger.info(f'''Param: {args}''' )
with open(os.path.join(args.dump_path,"""parameters.json""" ),"""w""" ) as f:
json.dump(vars(snake_case_ ),snake_case_,indent=4 )
git_log(args.dump_path )
_A , _A , _A : List[Any] = MODEL_CLASSES[args.student_type]
_A , _A , _A : Union[str, Any] = MODEL_CLASSES[args.teacher_type]
# TOKENIZER #
_A : Tuple = teacher_tokenizer_class.from_pretrained(args.teacher_name )
_A : List[Any] = {}
for tok_name, tok_symbol in tokenizer.special_tokens_map.items():
_A : List[Any] = tokenizer.all_special_tokens.index(snake_case_ )
_A : List[Any] = tokenizer.all_special_ids[idx]
logger.info(f'''Special tokens {special_tok_ids}''' )
_A : int = special_tok_ids
_A : Dict = tokenizer.max_model_input_sizes[args.teacher_name]
# DATA LOADER #
logger.info(f'''Loading data from {args.data_file}''' )
with open(args.data_file,"""rb""" ) as fp:
_A : List[Any] = pickle.load(snake_case_ )
if args.mlm:
logger.info(f'''Loading token counts from {args.token_counts} (already pre-computed)''' )
with open(args.token_counts,"""rb""" ) as fp:
_A : Optional[Any] = pickle.load(snake_case_ )
_A : Union[str, Any] = np.maximum(snake_case_,1 ) ** -args.mlm_smoothing
for idx in special_tok_ids.values():
_A : List[Any] = 0.0 # do not predict special tokens
_A : List[str] = torch.from_numpy(snake_case_ )
else:
_A : List[str] = None
_A : int = LmSeqsDataset(params=snake_case_,data=snake_case_ )
logger.info("""Data loader created.""" )
# STUDENT #
logger.info(f'''Loading student config from {args.student_config}''' )
_A : Tuple = student_config_class.from_pretrained(args.student_config )
_A : Dict = True
if args.student_pretrained_weights is not None:
logger.info(f'''Loading pretrained weights from {args.student_pretrained_weights}''' )
_A : List[Any] = student_model_class.from_pretrained(args.student_pretrained_weights,config=snake_case_ )
else:
_A : Optional[int] = student_model_class(snake_case_ )
if args.n_gpu > 0:
student.to(f'''cuda:{args.local_rank}''' )
logger.info("""Student loaded.""" )
# TEACHER #
_A : Any = teacher_model_class.from_pretrained(args.teacher_name,output_hidden_states=snake_case_ )
if args.n_gpu > 0:
teacher.to(f'''cuda:{args.local_rank}''' )
logger.info(f'''Teacher loaded from {args.teacher_name}.''' )
# FREEZING #
if args.freeze_pos_embs:
freeze_pos_embeddings(snake_case_,snake_case_ )
if args.freeze_token_type_embds:
freeze_token_type_embeddings(snake_case_,snake_case_ )
# SANITY CHECKS #
assert student.config.vocab_size == teacher.config.vocab_size
assert student.config.hidden_size == teacher.config.hidden_size
assert student.config.max_position_embeddings == teacher.config.max_position_embeddings
if args.mlm:
assert token_probs.size(0 ) == stu_architecture_config.vocab_size
# DISTILLER #
torch.cuda.empty_cache()
_A : Optional[int] = Distiller(
params=snake_case_,dataset=snake_case_,token_probs=snake_case_,student=snake_case_,teacher=snake_case_ )
distiller.train()
logger.info("""Let's go get some drinks.""" )
if __name__ == "__main__":
main()
| 26 |
import importlib
import json
import os
from collections import OrderedDict
from typing import Dict, Optional, Union
# Build the list of all feature extractors
from ...configuration_utils import PretrainedConfig
from ...dynamic_module_utils import get_class_from_dynamic_module, resolve_trust_remote_code
from ...feature_extraction_utils import FeatureExtractionMixin
from ...utils import CONFIG_NAME, FEATURE_EXTRACTOR_NAME, get_file_from_repo, logging
from .auto_factory import _LazyAutoMapping
from .configuration_auto import (
CONFIG_MAPPING_NAMES,
AutoConfig,
model_type_to_module_name,
replace_list_option_in_docstrings,
)
_snake_case = logging.get_logger(__name__)
_snake_case = OrderedDict(
[
("audio-spectrogram-transformer", "ASTFeatureExtractor"),
("beit", "BeitFeatureExtractor"),
("chinese_clip", "ChineseCLIPFeatureExtractor"),
("clap", "ClapFeatureExtractor"),
("clip", "CLIPFeatureExtractor"),
("clipseg", "ViTFeatureExtractor"),
("conditional_detr", "ConditionalDetrFeatureExtractor"),
("convnext", "ConvNextFeatureExtractor"),
("cvt", "ConvNextFeatureExtractor"),
("data2vec-audio", "Wav2Vec2FeatureExtractor"),
("data2vec-vision", "BeitFeatureExtractor"),
("deformable_detr", "DeformableDetrFeatureExtractor"),
("deit", "DeiTFeatureExtractor"),
("detr", "DetrFeatureExtractor"),
("dinat", "ViTFeatureExtractor"),
("donut-swin", "DonutFeatureExtractor"),
("dpt", "DPTFeatureExtractor"),
("encodec", "EncodecFeatureExtractor"),
("flava", "FlavaFeatureExtractor"),
("glpn", "GLPNFeatureExtractor"),
("groupvit", "CLIPFeatureExtractor"),
("hubert", "Wav2Vec2FeatureExtractor"),
("imagegpt", "ImageGPTFeatureExtractor"),
("layoutlmv2", "LayoutLMv2FeatureExtractor"),
("layoutlmv3", "LayoutLMv3FeatureExtractor"),
("levit", "LevitFeatureExtractor"),
("maskformer", "MaskFormerFeatureExtractor"),
("mctct", "MCTCTFeatureExtractor"),
("mobilenet_v1", "MobileNetV1FeatureExtractor"),
("mobilenet_v2", "MobileNetV2FeatureExtractor"),
("mobilevit", "MobileViTFeatureExtractor"),
("nat", "ViTFeatureExtractor"),
("owlvit", "OwlViTFeatureExtractor"),
("perceiver", "PerceiverFeatureExtractor"),
("poolformer", "PoolFormerFeatureExtractor"),
("regnet", "ConvNextFeatureExtractor"),
("resnet", "ConvNextFeatureExtractor"),
("segformer", "SegformerFeatureExtractor"),
("sew", "Wav2Vec2FeatureExtractor"),
("sew-d", "Wav2Vec2FeatureExtractor"),
("speech_to_text", "Speech2TextFeatureExtractor"),
("speecht5", "SpeechT5FeatureExtractor"),
("swiftformer", "ViTFeatureExtractor"),
("swin", "ViTFeatureExtractor"),
("swinv2", "ViTFeatureExtractor"),
("table-transformer", "DetrFeatureExtractor"),
("timesformer", "VideoMAEFeatureExtractor"),
("tvlt", "TvltFeatureExtractor"),
("unispeech", "Wav2Vec2FeatureExtractor"),
("unispeech-sat", "Wav2Vec2FeatureExtractor"),
("van", "ConvNextFeatureExtractor"),
("videomae", "VideoMAEFeatureExtractor"),
("vilt", "ViltFeatureExtractor"),
("vit", "ViTFeatureExtractor"),
("vit_mae", "ViTFeatureExtractor"),
("vit_msn", "ViTFeatureExtractor"),
("wav2vec2", "Wav2Vec2FeatureExtractor"),
("wav2vec2-conformer", "Wav2Vec2FeatureExtractor"),
("wavlm", "Wav2Vec2FeatureExtractor"),
("whisper", "WhisperFeatureExtractor"),
("xclip", "CLIPFeatureExtractor"),
("yolos", "YolosFeatureExtractor"),
]
)
_snake_case = _LazyAutoMapping(CONFIG_MAPPING_NAMES, FEATURE_EXTRACTOR_MAPPING_NAMES)
def lowerCAmelCase_ ( snake_case_ ):
for module_name, extractors in FEATURE_EXTRACTOR_MAPPING_NAMES.items():
if class_name in extractors:
_A : List[str] = model_type_to_module_name(snake_case_ )
_A : List[Any] = importlib.import_module(f'''.{module_name}''',"""transformers.models""" )
try:
return getattr(snake_case_,snake_case_ )
except AttributeError:
continue
for _, extractor in FEATURE_EXTRACTOR_MAPPING._extra_content.items():
if getattr(snake_case_,"""__name__""",snake_case_ ) == class_name:
return extractor
# We did not fine the class, but maybe it's because a dep is missing. In that case, the class will be in the main
# init and we return the proper dummy to get an appropriate error message.
_A : List[Any] = importlib.import_module("""transformers""" )
if hasattr(snake_case_,snake_case_ ):
return getattr(snake_case_,snake_case_ )
return None
def lowerCAmelCase_ ( snake_case_,snake_case_ = None,snake_case_ = False,snake_case_ = False,snake_case_ = None,snake_case_ = None,snake_case_ = None,snake_case_ = False,**snake_case_,):
_A : Optional[int] = get_file_from_repo(
snake_case_,snake_case_,cache_dir=snake_case_,force_download=snake_case_,resume_download=snake_case_,proxies=snake_case_,use_auth_token=snake_case_,revision=snake_case_,local_files_only=snake_case_,)
if resolved_config_file is None:
logger.info(
"""Could not locate the feature extractor configuration file, will try to use the model config instead.""" )
return {}
with open(snake_case_,encoding="""utf-8""" ) as reader:
return json.load(snake_case_ )
class lowercase :
def __init__( self ) -> List[Any]:
raise EnvironmentError(
"""AutoFeatureExtractor is designed to be instantiated """
"""using the `AutoFeatureExtractor.from_pretrained(pretrained_model_name_or_path)` method.""" )
@classmethod
@replace_list_option_in_docstrings(_a )
def a__ ( cls , _a , **_a ) -> Any:
_A : Tuple = kwargs.pop("""config""" , _a )
_A : Tuple = kwargs.pop("""trust_remote_code""" , _a )
_A : List[Any] = True
_A , _A : Tuple = FeatureExtractionMixin.get_feature_extractor_dict(_a , **_a )
_A : Tuple = config_dict.get("""feature_extractor_type""" , _a )
_A : int = None
if "AutoFeatureExtractor" in config_dict.get("""auto_map""" , {} ):
_A : Optional[int] = config_dict["""auto_map"""]["""AutoFeatureExtractor"""]
# If we don't find the feature extractor class in the feature extractor config, let's try the model config.
if feature_extractor_class is None and feature_extractor_auto_map is None:
if not isinstance(_a , _a ):
_A : int = AutoConfig.from_pretrained(_a , **_a )
# It could be in `config.feature_extractor_type``
_A : Optional[int] = getattr(_a , """feature_extractor_type""" , _a )
if hasattr(_a , """auto_map""" ) and "AutoFeatureExtractor" in config.auto_map:
_A : Tuple = config.auto_map["""AutoFeatureExtractor"""]
if feature_extractor_class is not None:
_A : Optional[Any] = feature_extractor_class_from_name(_a )
_A : List[Any] = feature_extractor_auto_map is not None
_A : Union[str, Any] = feature_extractor_class is not None or type(_a ) in FEATURE_EXTRACTOR_MAPPING
_A : Optional[int] = resolve_trust_remote_code(
_a , _a , _a , _a )
if has_remote_code and trust_remote_code:
_A : Dict = get_class_from_dynamic_module(
_a , _a , **_a )
_A : str = kwargs.pop("""code_revision""" , _a )
if os.path.isdir(_a ):
feature_extractor_class.register_for_auto_class()
return feature_extractor_class.from_dict(_a , **_a )
elif feature_extractor_class is not None:
return feature_extractor_class.from_dict(_a , **_a )
# Last try: we use the FEATURE_EXTRACTOR_MAPPING.
elif type(_a ) in FEATURE_EXTRACTOR_MAPPING:
_A : Dict = FEATURE_EXTRACTOR_MAPPING[type(_a )]
return feature_extractor_class.from_dict(_a , **_a )
raise ValueError(
F'''Unrecognized feature extractor in {pretrained_model_name_or_path}. Should have a '''
F'''`feature_extractor_type` key in its {FEATURE_EXTRACTOR_NAME} of {CONFIG_NAME}, or one of the following '''
F'''`model_type` keys in its {CONFIG_NAME}: {", ".join(c for c in FEATURE_EXTRACTOR_MAPPING_NAMES.keys() )}''' )
@staticmethod
def a__ ( _a , _a ) -> Optional[int]:
FEATURE_EXTRACTOR_MAPPING.register(_a , _a )
| 26 | 1 |
from __future__ import annotations
_snake_case = 1.6021e-19 # units = C
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,):
if (conductivity, electron_conc, mobility).count(0 ) != 1:
raise ValueError("""You cannot supply more or less than 2 values""" )
elif conductivity < 0:
raise ValueError("""Conductivity cannot be negative""" )
elif electron_conc < 0:
raise ValueError("""Electron concentration cannot be negative""" )
elif mobility < 0:
raise ValueError("""mobility cannot be negative""" )
elif conductivity == 0:
return (
"conductivity",
mobility * electron_conc * ELECTRON_CHARGE,
)
elif electron_conc == 0:
return (
"electron_conc",
conductivity / (mobility * ELECTRON_CHARGE),
)
else:
return (
"mobility",
conductivity / (electron_conc * ELECTRON_CHARGE),
)
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 |
import unittest
import numpy as np
from transformers.testing_utils import is_flaky, require_torch, require_vision
from transformers.utils import is_torch_available, is_vision_available
from ...test_image_processing_common import ImageProcessingSavingTestMixin, prepare_image_inputs
if is_torch_available():
import torch
if is_vision_available():
from PIL import Image
from transformers import DonutImageProcessor
class lowercase ( unittest.TestCase ):
def __init__( self , _a , _a=7 , _a=3 , _a=18 , _a=30 , _a=400 , _a=True , _a=None , _a=True , _a=False , _a=True , _a=True , _a=[0.5, 0.5, 0.5] , _a=[0.5, 0.5, 0.5] , ) -> Dict:
_A : str = parent
_A : int = batch_size
_A : Optional[int] = num_channels
_A : List[Any] = image_size
_A : int = min_resolution
_A : Optional[int] = max_resolution
_A : Any = do_resize
_A : List[str] = size if size is not None else {"""height""": 18, """width""": 20}
_A : Optional[int] = do_thumbnail
_A : str = do_align_axis
_A : List[Any] = do_pad
_A : Optional[Any] = do_normalize
_A : Tuple = image_mean
_A : List[str] = image_std
def a__ ( self ) -> Optional[int]:
return {
"do_resize": self.do_resize,
"size": self.size,
"do_thumbnail": self.do_thumbnail,
"do_align_long_axis": self.do_align_axis,
"do_pad": self.do_pad,
"do_normalize": self.do_normalize,
"image_mean": self.image_mean,
"image_std": self.image_std,
}
@require_torch
@require_vision
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = DonutImageProcessor if is_vision_available() else None
def a__ ( self ) -> Optional[int]:
_A : List[str] = DonutImageProcessingTester(self )
@property
def a__ ( self ) -> List[Any]:
return self.image_processor_tester.prepare_image_processor_dict()
def a__ ( self ) -> Optional[Any]:
_A : Union[str, Any] = self.image_processing_class(**self.image_processor_dict )
self.assertTrue(hasattr(_a , """do_resize""" ) )
self.assertTrue(hasattr(_a , """size""" ) )
self.assertTrue(hasattr(_a , """do_thumbnail""" ) )
self.assertTrue(hasattr(_a , """do_align_long_axis""" ) )
self.assertTrue(hasattr(_a , """do_pad""" ) )
self.assertTrue(hasattr(_a , """do_normalize""" ) )
self.assertTrue(hasattr(_a , """image_mean""" ) )
self.assertTrue(hasattr(_a , """image_std""" ) )
def a__ ( self ) -> List[Any]:
_A : Union[str, Any] = self.image_processing_class.from_dict(self.image_processor_dict )
self.assertEqual(image_processor.size , {"""height""": 18, """width""": 20} )
_A : Optional[int] = self.image_processing_class.from_dict(self.image_processor_dict , size=42 )
self.assertEqual(image_processor.size , {"""height""": 42, """width""": 42} )
# Previous config had dimensions in (width, height) order
_A : List[str] = self.image_processing_class.from_dict(self.image_processor_dict , size=(42, 84) )
self.assertEqual(image_processor.size , {"""height""": 84, """width""": 42} )
def a__ ( self ) -> Union[str, Any]:
pass
@is_flaky()
def a__ ( self ) -> Optional[int]:
# Initialize image_processing
_A : List[Any] = self.image_processing_class(**self.image_processor_dict )
# create random PIL images
_A : Dict = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a )
for image in image_inputs:
self.assertIsInstance(_a , Image.Image )
# Test not batched input
_A : Union[str, Any] = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : Any = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
@is_flaky()
def a__ ( self ) -> Dict:
# Initialize image_processing
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random numpy tensors
_A : Tuple = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , numpify=_a )
for image in image_inputs:
self.assertIsInstance(_a , np.ndarray )
# Test not batched input
_A : int = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : List[str] = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
@is_flaky()
def a__ ( self ) -> Optional[int]:
# Initialize image_processing
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random PyTorch tensors
_A : Optional[int] = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , torchify=_a )
for image in image_inputs:
self.assertIsInstance(_a , torch.Tensor )
# Test not batched input
_A : Any = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
1,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
# Test batched
_A : str = image_processing(_a , return_tensors="""pt""" ).pixel_values
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
self.image_processor_tester.size["""height"""],
self.image_processor_tester.size["""width"""],
) , )
| 26 | 1 |
import requests
from bsa import BeautifulSoup
def lowerCAmelCase_ ( snake_case_ = "AAPL" ):
_A : str = f'''https://in.finance.yahoo.com/quote/{symbol}?s={symbol}'''
_A : List[Any] = BeautifulSoup(requests.get(snake_case_ ).text,"""html.parser""" )
_A : Union[str, Any] = """My(6px) Pos(r) smartphone_Mt(6px)"""
return soup.find("""div""",class_=class_ ).find("""span""" ).text
if __name__ == "__main__":
for symbol in "AAPL AMZN IBM GOOG MSFT ORCL".split():
print(f"""Current {symbol:<4} stock price is {stock_price(symbol):>8}""")
| 26 |
from __future__ import annotations
import numpy as np
def lowerCAmelCase_ ( snake_case_ ):
return np.maximum(0,snake_case_ )
if __name__ == "__main__":
print(np.array(relu([-1, 0, 5]))) # --> [0, 0, 5]
| 26 | 1 |
import argparse
import numpy as np
import torch
from transformers import SpeechTaHifiGan, SpeechTaHifiGanConfig, logging
logging.set_verbosity_info()
_snake_case = logging.get_logger("transformers.models.speecht5")
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
hf_model.apply_weight_norm()
_A : Tuple = checkpoint["""input_conv.weight_g"""]
_A : List[Any] = checkpoint["""input_conv.weight_v"""]
_A : int = checkpoint["""input_conv.bias"""]
for i in range(len(config.upsample_rates ) ):
_A : Optional[Any] = checkpoint[f'''upsamples.{i}.1.weight_g''']
_A : Any = checkpoint[f'''upsamples.{i}.1.weight_v''']
_A : Optional[int] = checkpoint[f'''upsamples.{i}.1.bias''']
for i in range(len(config.upsample_rates ) * len(config.resblock_kernel_sizes ) ):
for j in range(len(config.resblock_dilation_sizes ) ):
_A : List[str] = checkpoint[f'''blocks.{i}.convs1.{j}.1.weight_g''']
_A : Optional[int] = checkpoint[f'''blocks.{i}.convs1.{j}.1.weight_v''']
_A : List[Any] = checkpoint[f'''blocks.{i}.convs1.{j}.1.bias''']
_A : Tuple = checkpoint[f'''blocks.{i}.convs2.{j}.1.weight_g''']
_A : Union[str, Any] = checkpoint[f'''blocks.{i}.convs2.{j}.1.weight_v''']
_A : Union[str, Any] = checkpoint[f'''blocks.{i}.convs2.{j}.1.bias''']
_A : Optional[int] = checkpoint["""output_conv.1.weight_g"""]
_A : str = checkpoint["""output_conv.1.weight_v"""]
_A : Union[str, Any] = checkpoint["""output_conv.1.bias"""]
hf_model.remove_weight_norm()
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_=None,snake_case_=None,):
if config_path is not None:
_A : List[Any] = SpeechTaHifiGanConfig.from_pretrained(snake_case_ )
else:
_A : int = SpeechTaHifiGanConfig()
_A : List[Any] = SpeechTaHifiGan(snake_case_ )
_A : Optional[Any] = torch.load(snake_case_ )
load_weights(orig_checkpoint["""model"""]["""generator"""],snake_case_,snake_case_ )
_A : Tuple = np.load(snake_case_ )
_A : Dict = stats[0].reshape(-1 )
_A : List[Any] = stats[1].reshape(-1 )
_A : Tuple = torch.from_numpy(snake_case_ ).float()
_A : str = torch.from_numpy(snake_case_ ).float()
model.save_pretrained(snake_case_ )
if repo_id:
print("""Pushing to the hub...""" )
model.push_to_hub(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
parser.add_argument("--checkpoint_path", required=True, default=None, type=str, help="Path to original checkpoint")
parser.add_argument("--stats_path", required=True, default=None, type=str, help="Path to stats.npy file")
parser.add_argument("--config_path", default=None, type=str, help="Path to hf config.json of model to convert")
parser.add_argument(
"--pytorch_dump_folder_path", required=True, default=None, type=str, help="Path to the output PyTorch model."
)
parser.add_argument(
"--push_to_hub", default=None, type=str, help="Where to upload the converted model on the 🤗 hub."
)
_snake_case = parser.parse_args()
convert_hifigan_checkpoint(
args.checkpoint_path,
args.stats_path,
args.pytorch_dump_folder_path,
args.config_path,
args.push_to_hub,
)
| 26 |
import argparse
import shutil
import time
from json import JSONDecodeError
from logging import getLogger
from pathlib import Path
from typing import Dict, List
import torch
from torch.utils.data import DataLoader
from tqdm import tqdm
from transformers import AutoModelForSeqaSeqLM, AutoTokenizer
from utils import (
SeqaSeqDataset,
calculate_bleu,
calculate_rouge,
chunks,
lmap,
load_json,
parse_numeric_n_bool_cl_kwargs,
save_json,
use_task_specific_params,
write_txt_file,
)
_snake_case = getLogger(__name__)
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ = 8,snake_case_ = 1024,snake_case_="val",snake_case_=None,snake_case_=False,snake_case_="summarization",snake_case_=None,snake_case_=1,snake_case_ = None,snake_case_="",**snake_case_,):
_A : Dict = str(snake_case_ )
assert local_rank is not None
torch.distributed.init_process_group(backend="""nccl""",rank=snake_case_ )
_A : Tuple = Path(snake_case_ )
_A : List[Any] = save_dir.joinpath(f'''rank_{local_rank}_output.json''' )
torch.cuda.set_device(snake_case_ )
_A : Union[str, Any] = AutoModelForSeqaSeqLM.from_pretrained(snake_case_ ).cuda()
if fpaa:
_A : Any = model.half()
# determine if we need to increase num_beams
use_task_specific_params(snake_case_,snake_case_ ) # update config with task specific params
_A : str = generate_kwargs.pop("""num_beams""",model.config.num_beams ) # AttributeError risk?
if num_return_sequences > num_beams:
_A : int = num_return_sequences
_A : Optional[Any] = AutoTokenizer.from_pretrained(snake_case_ )
logger.info(f'''Inferred tokenizer type: {tokenizer.__class__}''' ) # if this is wrong, check config.model_type.
if max_source_length is None:
_A : Optional[int] = tokenizer.model_max_length
if prefix is None:
_A : Tuple = prefix or getattr(model.config,"""prefix""","""""" ) or """"""
_A : Optional[int] = SeqaSeqDataset(
snake_case_,snake_case_,snake_case_,max_target_length=1024,type_path=snake_case_,n_obs=snake_case_,prefix=snake_case_,**snake_case_,)
# I set shuffle=True for a more accurate progress bar.
# If all the longest samples are first, the prog bar estimate is too high at the beginning.
_A : Optional[int] = ds.make_sortish_sampler(snake_case_,distributed=snake_case_,add_extra_examples=snake_case_,shuffle=snake_case_ )
_A : Dict = DataLoader(snake_case_,sampler=snake_case_,batch_size=snake_case_,collate_fn=ds.collate_fn )
_A : Optional[Any] = []
for batch in tqdm(snake_case_ ):
_A : Tuple = model.generate(
input_ids=batch["""input_ids"""].to(model.device ),attention_mask=batch["""attention_mask"""].to(model.device ),num_return_sequences=snake_case_,num_beams=snake_case_,**snake_case_,)
_A : Any = tokenizer.batch_decode(snake_case_,skip_special_tokens=snake_case_,clean_up_tokenization_spaces=snake_case_ )
_A : Dict = batch["""ids"""]
if num_return_sequences > 1:
_A : Any = chunks(snake_case_,snake_case_ ) # batch size chunks, each of size num_return_seq
for i, pred in enumerate(snake_case_ ):
results.append({"""pred""": pred, """id""": ids[i].item()} )
save_json(snake_case_,snake_case_ )
return results, sampler.num_replicas
def lowerCAmelCase_ ( ):
_A : Tuple = argparse.ArgumentParser(
epilog="""Unspecified args like --num_beams=2 --decoder_start_token_id=4 are passed to model.generate""" )
parser.add_argument("""--data_dir""",type=snake_case_,help="""like cnn_dm/test.source""" )
parser.add_argument(
"""--model_name""",type=snake_case_,help="""like facebook/bart-large-cnn,t5-base, etc.""",default="""sshleifer/distilbart-xsum-12-3""",)
parser.add_argument("""--save_dir""",type=snake_case_,help="""where to save""",default="""tmp_gen""" )
parser.add_argument("""--max_source_length""",type=snake_case_,default=snake_case_ )
parser.add_argument(
"""--type_path""",type=snake_case_,default="""test""",help="""which subset to evaluate typically train/val/test""" )
parser.add_argument("""--task""",type=snake_case_,default="""summarization""",help="""used for task_specific_params + metrics""" )
parser.add_argument("""--bs""",type=snake_case_,default=8,required=snake_case_,help="""batch size""" )
parser.add_argument(
"""--local_rank""",type=snake_case_,default=-1,required=snake_case_,help="""should be passed by distributed.launch""" )
parser.add_argument(
"""--n_obs""",type=snake_case_,default=snake_case_,required=snake_case_,help="""How many observations. Defaults to all.""" )
parser.add_argument(
"""--num_return_sequences""",type=snake_case_,default=1,required=snake_case_,help="""How many sequences to return""" )
parser.add_argument(
"""--sync_timeout""",type=snake_case_,default=600,required=snake_case_,help="""How long should master process wait for other processes to finish.""",)
parser.add_argument("""--src_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument("""--tgt_lang""",type=snake_case_,default=snake_case_,required=snake_case_ )
parser.add_argument(
"""--prefix""",type=snake_case_,required=snake_case_,default=snake_case_,help="""will be added to the begininng of src examples""" )
parser.add_argument("""--fp16""",action="""store_true""" )
parser.add_argument("""--debug""",action="""store_true""" )
_A : Union[str, Any] = time.time()
_A , _A : List[str] = parser.parse_known_args()
_A : List[str] = parse_numeric_n_bool_cl_kwargs(snake_case_ )
if generate_kwargs and args.local_rank <= 0:
print(f'''parsed the following generate kwargs: {generate_kwargs}''' )
_A : Dict = Path(args.save_dir + """_tmp""" )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ ) # this handles locking.
_A : int = list(json_save_dir.glob("""rank_*.json""" ) )
if intermediate_files:
raise ValueError(f'''Found files at {json_save_dir} please move or remove them.''' )
# In theory, a node could finish and save before another node hits this. If this happens, we can address later.
_A : Any = {}
if args.src_lang is not None:
_A : int = args.src_lang
if args.tgt_lang is not None:
_A : Dict = args.tgt_lang
Path(args.save_dir ).mkdir(exist_ok=snake_case_ )
_A , _A : str = eval_data_dir(
args.data_dir,snake_case_,args.model_name,type_path=args.type_path,bs=args.bs,fpaa=args.fpaa,task=args.task,local_rank=args.local_rank,n_obs=args.n_obs,max_source_length=args.max_source_length,num_return_sequences=args.num_return_sequences,prefix=args.prefix,dataset_kwargs=snake_case_,**snake_case_,)
if args.local_rank <= 0:
_A : List[Any] = Path(args.save_dir )
save_dir.mkdir(exist_ok=snake_case_ )
_A : Tuple = gather_results_from_each_node(snake_case_,snake_case_,args.sync_timeout )
_A : Optional[int] = combine_partial_results(snake_case_ )
if args.num_return_sequences > 1:
_A : Optional[Any] = save_dir.joinpath("""pseudolabel_results.json""" )
print(f'''Saving aggregated results at {save_path}, intermediate in {json_save_dir}/''' )
save_json(snake_case_,snake_case_ )
return
_A : List[str] = Path(args.data_dir ).joinpath(args.type_path + """.target""" )
with open(snake_case_ ) as f:
_A : int = [x.rstrip() for x in f.readlines()][: len(snake_case_ )]
# Calculate metrics, save metrics, and save _generations.txt
_A : Dict = """translation""" in args.task
_A : Optional[Any] = calculate_bleu if calc_bleu else calculate_rouge
_A : Tuple = """bleu""" if calc_bleu else """rouge"""
_A : Dict = score_fn(snake_case_,snake_case_ )
_A : List[Any] = len(snake_case_ )
_A : Optional[int] = time.time() - start_time
_A : Dict = round(runtime / metrics["""n_obs"""],4 )
_A : Dict = num_replicas
# TODO(@stas00): add whatever metadata to metrics
_A : Any = save_dir.joinpath(f'''{args.type_path}_{metric_name}.json''' )
save_json(snake_case_,snake_case_,indent=snake_case_ )
print(snake_case_ )
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}_generations.txt''' ) )
if args.debug:
write_txt_file(snake_case_,save_dir.joinpath(f'''{args.type_path}.target''' ) )
else:
shutil.rmtree(snake_case_ )
def lowerCAmelCase_ ( snake_case_ ):
_A : Dict = []
for partial_result in partial_results:
records.extend(snake_case_ )
_A : Optional[Any] = sorted(snake_case_,key=lambda snake_case_ : x["id"] )
_A : List[str] = [x["""pred"""] for x in records]
return preds
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
# WAIT FOR lots of .json files
_A : Optional[Any] = time.time()
logger.info("""waiting for all nodes to finish""" )
_A : List[str] = None
while (time.time() - start_wait) < timeout:
_A : str = list(save_dir.glob("""rank_*.json""" ) )
if len(snake_case_ ) < num_replicas:
continue
try:
# make sure all json files are fully saved
_A : List[str] = lmap(snake_case_,snake_case_ )
return json_data
except JSONDecodeError:
continue
else:
raise TimeoutError("""Rank 0 gave up on waiting for other processes""" )
# Unreachable
if __name__ == "__main__":
# Usage for MT:
run_generate()
| 26 | 1 |
import json
from typing import List, Optional, Tuple
from tokenizers import normalizers
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import logging
from .tokenization_distilbert import DistilBertTokenizer
_snake_case = logging.get_logger(__name__)
_snake_case = {"vocab_file": "vocab.txt", "tokenizer_file": "tokenizer.json"}
_snake_case = {
"vocab_file": {
"distilbert-base-uncased": "https://huggingface.co/distilbert-base-uncased/resolve/main/vocab.txt",
"distilbert-base-uncased-distilled-squad": (
"https://huggingface.co/distilbert-base-uncased-distilled-squad/resolve/main/vocab.txt"
),
"distilbert-base-cased": "https://huggingface.co/distilbert-base-cased/resolve/main/vocab.txt",
"distilbert-base-cased-distilled-squad": (
"https://huggingface.co/distilbert-base-cased-distilled-squad/resolve/main/vocab.txt"
),
"distilbert-base-german-cased": "https://huggingface.co/distilbert-base-german-cased/resolve/main/vocab.txt",
"distilbert-base-multilingual-cased": (
"https://huggingface.co/distilbert-base-multilingual-cased/resolve/main/vocab.txt"
),
},
"tokenizer_file": {
"distilbert-base-uncased": "https://huggingface.co/distilbert-base-uncased/resolve/main/tokenizer.json",
"distilbert-base-uncased-distilled-squad": (
"https://huggingface.co/distilbert-base-uncased-distilled-squad/resolve/main/tokenizer.json"
),
"distilbert-base-cased": "https://huggingface.co/distilbert-base-cased/resolve/main/tokenizer.json",
"distilbert-base-cased-distilled-squad": (
"https://huggingface.co/distilbert-base-cased-distilled-squad/resolve/main/tokenizer.json"
),
"distilbert-base-german-cased": (
"https://huggingface.co/distilbert-base-german-cased/resolve/main/tokenizer.json"
),
"distilbert-base-multilingual-cased": (
"https://huggingface.co/distilbert-base-multilingual-cased/resolve/main/tokenizer.json"
),
},
}
_snake_case = {
"distilbert-base-uncased": 512,
"distilbert-base-uncased-distilled-squad": 512,
"distilbert-base-cased": 512,
"distilbert-base-cased-distilled-squad": 512,
"distilbert-base-german-cased": 512,
"distilbert-base-multilingual-cased": 512,
}
_snake_case = {
"distilbert-base-uncased": {"do_lower_case": True},
"distilbert-base-uncased-distilled-squad": {"do_lower_case": True},
"distilbert-base-cased": {"do_lower_case": False},
"distilbert-base-cased-distilled-squad": {"do_lower_case": False},
"distilbert-base-german-cased": {"do_lower_case": False},
"distilbert-base-multilingual-cased": {"do_lower_case": False},
}
class lowercase ( UpperCamelCase__ ):
_a = VOCAB_FILES_NAMES
_a = PRETRAINED_VOCAB_FILES_MAP
_a = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
_a = PRETRAINED_INIT_CONFIGURATION
_a = ["input_ids", "attention_mask"]
_a = DistilBertTokenizer
def __init__( self , _a=None , _a=None , _a=True , _a="[UNK]" , _a="[SEP]" , _a="[PAD]" , _a="[CLS]" , _a="[MASK]" , _a=True , _a=None , **_a , ) -> Tuple:
super().__init__(
_a , tokenizer_file=_a , do_lower_case=_a , unk_token=_a , sep_token=_a , pad_token=_a , cls_token=_a , mask_token=_a , tokenize_chinese_chars=_a , strip_accents=_a , **_a , )
_A : List[Any] = json.loads(self.backend_tokenizer.normalizer.__getstate__() )
if (
normalizer_state.get("""lowercase""" , _a ) != do_lower_case
or normalizer_state.get("""strip_accents""" , _a ) != strip_accents
or normalizer_state.get("""handle_chinese_chars""" , _a ) != tokenize_chinese_chars
):
_A : str = getattr(_a , normalizer_state.pop("""type""" ) )
_A : str = do_lower_case
_A : Any = strip_accents
_A : Optional[int] = tokenize_chinese_chars
_A : int = normalizer_class(**_a )
_A : str = do_lower_case
def a__ ( self , _a , _a=None ) -> Dict:
_A : Tuple = [self.cls_token_id] + token_ids_a + [self.sep_token_id]
if token_ids_a:
output += token_ids_a + [self.sep_token_id]
return output
def a__ ( self , _a , _a = None ) -> List[int]:
_A : Dict = [self.sep_token_id]
_A : Optional[Any] = [self.cls_token_id]
if token_ids_a is None:
return len(cls + token_ids_a + sep ) * [0]
return len(cls + token_ids_a + sep ) * [0] + len(token_ids_a + sep ) * [1]
def a__ ( self , _a , _a = None ) -> Tuple[str]:
_A : Optional[Any] = self._tokenizer.model.save(_a , name=_a )
return tuple(_a )
| 26 |
from __future__ import annotations
import unittest
from transformers import is_tf_available
from transformers.testing_utils import require_sentencepiece, require_tf, require_tokenizers, slow
if is_tf_available():
import numpy as np
import tensorflow as tf
from transformers import TFCamembertModel
@require_tf
@require_sentencepiece
@require_tokenizers
class lowercase ( unittest.TestCase ):
@slow
def a__ ( self ) -> Any:
_A : Tuple = TFCamembertModel.from_pretrained("""jplu/tf-camembert-base""" )
_A : List[Any] = tf.convert_to_tensor(
[[5, 121, 11, 660, 16, 730, 2_5543, 110, 83, 6]] , dtype=tf.intaa , ) # J'aime le camembert !"
_A : List[str] = model(_a )["""last_hidden_state"""]
_A : Union[str, Any] = tf.TensorShape((1, 10, 768) )
self.assertEqual(output.shape , _a )
# compare the actual values for a slice.
_A : List[Any] = tf.convert_to_tensor(
[[[-0.0254, 0.0235, 0.1027], [0.0606, -0.1811, -0.0418], [-0.1561, -0.1127, 0.2687]]] , dtype=tf.floataa , )
# camembert = torch.hub.load('pytorch/fairseq', 'camembert.v0')
# camembert.eval()
# expected_slice = roberta.model.forward(input_ids)[0][:, :3, :3].detach()
self.assertTrue(np.allclose(output[:, :3, :3].numpy() , expected_slice.numpy() , atol=1e-4 ) )
| 26 | 1 |
import json
import multiprocessing as mp
import re
from collections import defaultdict
from functools import partial
from typing import Dict, List, Optional, Set, Tuple, Type
from datasets import Dataset
from datasketch import MinHash, MinHashLSH
from dpu_utils.utils.iterators import ThreadedIterator
from tqdm import tqdm
_snake_case = re.compile("[^A-Za-z_0-9]")
# parameters used in DuplicationIndex
_snake_case = 10
_snake_case = 256
def lowerCAmelCase_ ( snake_case_ ):
if len(snake_case_ ) < MIN_NUM_TOKENS:
return None
_A : Tuple = MinHash(num_perm=snake_case_ )
for token in set(snake_case_ ):
min_hash.update(token.encode() )
return min_hash
def lowerCAmelCase_ ( snake_case_ ):
return {t for t in NON_ALPHA.split(snake_case_ ) if len(t.strip() ) > 0}
class lowercase :
def __init__( self , *,
_a = 0.85 , ) -> List[Any]:
_A : Any = duplication_jaccard_threshold
_A : int = NUM_PERM
_A : List[str] = MinHashLSH(threshold=self._duplication_jaccard_threshold , num_perm=self._num_perm )
_A : str = defaultdict(_a )
def a__ ( self , _a , _a ) -> None:
_A : Any = self._index.query(_a )
if code_key in self._index.keys:
print(F'''Duplicate key {code_key}''' )
return
self._index.insert(_a , _a )
if len(_a ) > 0:
for base_duplicate in close_duplicates:
if base_duplicate in self._duplicate_clusters:
self._duplicate_clusters[base_duplicate].add(_a )
break
else:
self._duplicate_clusters[close_duplicates[0]].add(_a )
def a__ ( self ) -> List[List[Dict]]:
_A : List[Any] = []
for base, duplicates in self._duplicate_clusters.items():
_A : Dict = [base] + list(_a )
# reformat the cluster to be a list of dict
_A : Tuple = [{"""base_index""": el[0], """repo_name""": el[1], """path""": el[2]} for el in cluster]
duplicate_clusters.append(_a )
return duplicate_clusters
def a__ ( self , _a ) -> None:
_A : str = self.get_duplicate_clusters()
with open(_a , """w""" ) as f:
json.dump(_a , _a )
def lowerCAmelCase_ ( snake_case_ ):
_A , _A : Tuple = element
_A : Any = get_min_hash([t for t in NON_ALPHA.split(data["""content"""] ) if len(t.strip() ) > 0] )
if min_hash is not None:
return (index, data["repo_name"], data["path"]), min_hash
def lowerCAmelCase_ ( snake_case_ ):
with mp.Pool() as pool:
for data in pool.imap_unordered(
_compute_min_hash,ThreadedIterator(snake_case_,max_queue_size=10000 ),chunksize=100,):
if data is not None:
yield data
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : int = DuplicationIndex(duplication_jaccard_threshold=snake_case_ )
for filename, min_hash in tqdm(ThreadedIterator(minhash_iter(enumerate(snake_case_ ) ),max_queue_size=100 ) ):
di.add(snake_case_,snake_case_ )
# Returns a List[Cluster] where Cluster is List[str] with the filenames.
return di.get_duplicate_clusters()
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Any = get_tokens(snake_case_ )
_A : Dict = get_tokens(snake_case_ )
return len(tokensa & tokensa ) / len(tokensa | tokensa )
_snake_case = None
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Optional[Any] = []
for elementa in cluster:
_A : List[str] = _shared_dataset[elementa["""base_index"""]]["""content"""]
for elementa in extremes:
_A : Union[str, Any] = _shared_dataset[elementa["""base_index"""]]["""content"""]
if jaccard_similarity(snake_case_,snake_case_ ) >= jaccard_threshold:
elementa["copies"] += 1
break
else:
_A : int = 1
extremes.append(snake_case_ )
return extremes
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
global _shared_dataset
_A : Any = dataset
_A : List[Any] = []
_A : Optional[int] = partial(_find_cluster_extremes_shared,jaccard_threshold=snake_case_ )
with mp.Pool() as pool:
for extremes in tqdm(
pool.imap_unordered(
snake_case_,snake_case_,),total=len(snake_case_ ),):
extremes_list.append(snake_case_ )
return extremes_list
def lowerCAmelCase_ ( snake_case_,snake_case_ = 0.85 ):
_A : List[Any] = make_duplicate_clusters(snake_case_,snake_case_ )
_A : str = {x["""base_index"""] for cluster in duplicate_clusters for x in cluster}
_A : List[str] = {}
_A : Optional[Any] = find_extremes(snake_case_,snake_case_,snake_case_ )
for extremes in extremes_clusters:
for element in extremes:
_A : int = element
_A : Union[str, Any] = duplicate_indices - set(extreme_dict.keys() )
_A : int = dataset.filter(lambda snake_case_,snake_case_ : idx not in remove_indices,with_indices=snake_case_ )
# update duplicate_clusters
for cluster in duplicate_clusters:
for element in cluster:
_A : int = element["""base_index"""] in extreme_dict
if element["is_extreme"]:
_A : Any = extreme_dict[element["""base_index"""]]["""copies"""]
print(f'''Original dataset size: {len(snake_case_ )}''' )
print(f'''Number of duplicate clusters: {len(snake_case_ )}''' )
print(f'''Files in duplicate cluster: {len(snake_case_ )}''' )
print(f'''Unique files in duplicate cluster: {len(snake_case_ )}''' )
print(f'''Filtered dataset size: {len(snake_case_ )}''' )
return ds_filter, duplicate_clusters
| 26 |
from dataclasses import dataclass
from typing import List, Optional, Union
import numpy as np
import PIL
import torch
from transformers import CLIPImageProcessor, CLIPVisionModel
from ...models import PriorTransformer
from ...pipelines import DiffusionPipeline
from ...schedulers import HeunDiscreteScheduler
from ...utils import (
BaseOutput,
is_accelerate_available,
logging,
randn_tensor,
replace_example_docstring,
)
from .renderer import ShapERenderer
_snake_case = logging.get_logger(__name__) # pylint: disable=invalid-name
_snake_case = "\n Examples:\n ```py\n >>> from PIL import Image\n >>> import torch\n >>> from diffusers import DiffusionPipeline\n >>> from diffusers.utils import export_to_gif, load_image\n\n >>> device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n\n >>> repo = \"openai/shap-e-img2img\"\n >>> pipe = DiffusionPipeline.from_pretrained(repo, torch_dtype=torch.float16)\n >>> pipe = pipe.to(device)\n\n >>> guidance_scale = 3.0\n >>> image_url = \"https://hf.co/datasets/diffusers/docs-images/resolve/main/shap-e/corgi.png\"\n >>> image = load_image(image_url).convert(\"RGB\")\n\n >>> images = pipe(\n ... image,\n ... guidance_scale=guidance_scale,\n ... num_inference_steps=64,\n ... frame_size=256,\n ... ).images\n\n >>> gif_path = export_to_gif(images[0], \"corgi_3d.gif\")\n ```\n"
@dataclass
class lowercase ( UpperCamelCase__ ):
_a = 42
class lowercase ( UpperCamelCase__ ):
def __init__( self , _a , _a , _a , _a , _a , ) -> List[Any]:
super().__init__()
self.register_modules(
prior=_a , image_encoder=_a , image_processor=_a , scheduler=_a , renderer=_a , )
def a__ ( self , _a , _a , _a , _a , _a , _a ) -> str:
if latents is None:
_A : str = randn_tensor(_a , generator=_a , device=_a , dtype=_a )
else:
if latents.shape != shape:
raise ValueError(F'''Unexpected latents shape, got {latents.shape}, expected {shape}''' )
_A : Union[str, Any] = latents.to(_a )
_A : int = latents * scheduler.init_noise_sigma
return latents
def a__ ( self , _a=0 ) -> Optional[Any]:
if is_accelerate_available():
from accelerate import cpu_offload
else:
raise ImportError("""Please install accelerate via `pip install accelerate`""" )
_A : str = torch.device(F'''cuda:{gpu_id}''' )
_A : Any = [self.image_encoder, self.prior]
for cpu_offloaded_model in models:
if cpu_offloaded_model is not None:
cpu_offload(_a , _a )
@property
def a__ ( self ) -> List[Any]:
if self.device != torch.device("""meta""" ) or not hasattr(self.image_encoder , """_hf_hook""" ):
return self.device
for module in self.image_encoder.modules():
if (
hasattr(_a , """_hf_hook""" )
and hasattr(module._hf_hook , """execution_device""" )
and module._hf_hook.execution_device is not None
):
return torch.device(module._hf_hook.execution_device )
return self.device
def a__ ( self , _a , _a , _a , _a , ) -> Tuple:
if isinstance(_a , _a ) and isinstance(image[0] , torch.Tensor ):
_A : int = torch.cat(_a , axis=0 ) if image[0].ndim == 4 else torch.stack(_a , axis=0 )
if not isinstance(_a , torch.Tensor ):
_A : Dict = self.image_processor(_a , return_tensors="""pt""" ).pixel_values[0].unsqueeze(0 )
_A : int = image.to(dtype=self.image_encoder.dtype , device=_a )
_A : List[Any] = self.image_encoder(_a )["""last_hidden_state"""]
_A : List[Any] = image_embeds[:, 1:, :].contiguous() # batch_size, dim, 256
_A : Dict = image_embeds.repeat_interleave(_a , dim=0 )
if do_classifier_free_guidance:
_A : str = torch.zeros_like(_a )
# For classifier free guidance, we need to do two forward passes.
# Here we concatenate the unconditional and text embeddings into a single batch
# to avoid doing two forward passes
_A : List[str] = torch.cat([negative_image_embeds, image_embeds] )
return image_embeds
@torch.no_grad()
@replace_example_docstring(_a )
def __call__( self , _a , _a = 1 , _a = 25 , _a = None , _a = None , _a = 4.0 , _a = 64 , _a = "pil" , _a = True , ) -> Union[str, Any]:
if isinstance(_a , PIL.Image.Image ):
_A : List[Any] = 1
elif isinstance(_a , torch.Tensor ):
_A : Any = image.shape[0]
elif isinstance(_a , _a ) and isinstance(image[0] , (torch.Tensor, PIL.Image.Image) ):
_A : Union[str, Any] = len(_a )
else:
raise ValueError(
F'''`image` has to be of type `PIL.Image.Image`, `torch.Tensor`, `List[PIL.Image.Image]` or `List[torch.Tensor]` but is {type(_a )}''' )
_A : Optional[int] = self._execution_device
_A : Tuple = batch_size * num_images_per_prompt
_A : List[Any] = guidance_scale > 1.0
_A : Optional[Any] = self._encode_image(_a , _a , _a , _a )
# prior
self.scheduler.set_timesteps(_a , device=_a )
_A : Optional[int] = self.scheduler.timesteps
_A : List[str] = self.prior.config.num_embeddings
_A : int = self.prior.config.embedding_dim
_A : Optional[Any] = self.prepare_latents(
(batch_size, num_embeddings * embedding_dim) , image_embeds.dtype , _a , _a , _a , self.scheduler , )
# YiYi notes: for testing only to match ldm, we can directly create a latents with desired shape: batch_size, num_embeddings, embedding_dim
_A : List[Any] = latents.reshape(latents.shape[0] , _a , _a )
for i, t in enumerate(self.progress_bar(_a ) ):
# expand the latents if we are doing classifier free guidance
_A : List[str] = torch.cat([latents] * 2 ) if do_classifier_free_guidance else latents
_A : int = self.scheduler.scale_model_input(_a , _a )
_A : Tuple = self.prior(
_a , timestep=_a , proj_embedding=_a , ).predicted_image_embedding
# remove the variance
_A , _A : Optional[Any] = noise_pred.split(
scaled_model_input.shape[2] , dim=2 ) # batch_size, num_embeddings, embedding_dim
if do_classifier_free_guidance is not None:
_A , _A : Dict = noise_pred.chunk(2 )
_A : Tuple = noise_pred_uncond + guidance_scale * (noise_pred - noise_pred_uncond)
_A : int = self.scheduler.step(
_a , timestep=_a , sample=_a , ).prev_sample
if output_type == "latent":
return ShapEPipelineOutput(images=_a )
_A : List[str] = []
for i, latent in enumerate(_a ):
print()
_A : List[str] = self.renderer.decode(
latent[None, :] , _a , size=_a , ray_batch_size=4096 , n_coarse_samples=64 , n_fine_samples=128 , )
images.append(_a )
_A : List[Any] = torch.stack(_a )
if output_type not in ["np", "pil"]:
raise ValueError(F'''Only the output types `pil` and `np` are supported not output_type={output_type}''' )
_A : List[str] = images.cpu().numpy()
if output_type == "pil":
_A : List[Any] = [self.numpy_to_pil(_a ) for image in images]
# Offload last model to CPU
if hasattr(self , """final_offload_hook""" ) and self.final_offload_hook is not None:
self.final_offload_hook.offload()
if not return_dict:
return (images,)
return ShapEPipelineOutput(images=_a )
| 26 | 1 |
from ...configuration_utils import PretrainedConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/config.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/config.json"
# See all FNet models at https://huggingface.co/models?filter=fnet
}
class lowercase ( UpperCamelCase__ ):
_a = "fnet"
def __init__( self , _a=3_2000 , _a=768 , _a=12 , _a=3072 , _a="gelu_new" , _a=0.1 , _a=512 , _a=4 , _a=0.02 , _a=1e-12 , _a=False , _a=512 , _a=3 , _a=1 , _a=2 , **_a , ) -> int:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Any = vocab_size
_A : str = max_position_embeddings
_A : Optional[Any] = hidden_size
_A : List[str] = num_hidden_layers
_A : List[str] = intermediate_size
_A : List[Any] = hidden_act
_A : List[str] = hidden_dropout_prob
_A : List[str] = initializer_range
_A : List[Any] = type_vocab_size
_A : List[Any] = layer_norm_eps
_A : List[str] = use_tpu_fourier_optimizations
_A : str = tpu_short_seq_length
| 26 |
import argparse
import collections
import json
from pathlib import Path
import requests
import torch
import yaml
from huggingface_hub import hf_hub_download
from PIL import Image
from transformers import (
MobileViTImageProcessor,
MobileViTVaConfig,
MobileViTVaForImageClassification,
MobileViTVaForSemanticSegmentation,
)
from transformers.utils import logging
logging.set_verbosity_info()
_snake_case = logging.get_logger(__name__)
def lowerCAmelCase_ ( snake_case_ ):
print("""Loading config file...""" )
def flatten_yaml_as_dict(snake_case_,snake_case_="",snake_case_="." ):
_A : Union[str, Any] = []
for k, v in d.items():
_A : Optional[int] = parent_key + sep + k if parent_key else k
if isinstance(snake_case_,collections.abc.MutableMapping ):
items.extend(flatten_yaml_as_dict(snake_case_,snake_case_,sep=snake_case_ ).items() )
else:
items.append((new_key, v) )
return dict(snake_case_ )
_A : List[Any] = argparse.Namespace()
with open(snake_case_,"""r""" ) as yaml_file:
try:
_A : List[Any] = yaml.load(snake_case_,Loader=yaml.FullLoader )
_A : Optional[int] = flatten_yaml_as_dict(snake_case_ )
for k, v in flat_cfg.items():
setattr(snake_case_,snake_case_,snake_case_ )
except yaml.YAMLError as exc:
logger.error("""Error while loading config file: {}. Error message: {}""".format(snake_case_,str(snake_case_ ) ) )
return config
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Optional[Any] = MobileViTVaConfig()
_A : Tuple = False
# dataset
if task_name.startswith("""imagenet1k_""" ):
_A : Dict = 1000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : int = 384
else:
_A : int = 256
_A : List[str] = """imagenet-1k-id2label.json"""
elif task_name.startswith("""imagenet21k_to_1k_""" ):
_A : Union[str, Any] = 21000
if int(task_name.strip().split("""_""" )[-1] ) == 384:
_A : str = 384
else:
_A : List[Any] = 256
_A : List[str] = """imagenet-22k-id2label.json"""
elif task_name.startswith("""ade20k_""" ):
_A : int = 151
_A : int = 512
_A : Optional[int] = """ade20k-id2label.json"""
_A : Any = True
elif task_name.startswith("""voc_""" ):
_A : List[Any] = 21
_A : Dict = 512
_A : Dict = """pascal-voc-id2label.json"""
_A : int = True
# orig_config
_A : Any = load_orig_config_file(snake_case_ )
assert getattr(snake_case_,"""model.classification.name""",-1 ) == "mobilevit_v2", "Invalid model"
_A : List[Any] = getattr(snake_case_,"""model.classification.mitv2.width_multiplier""",1.0 )
assert (
getattr(snake_case_,"""model.classification.mitv2.attn_norm_layer""",-1 ) == "layer_norm_2d"
), "Norm layers other than layer_norm_2d is not supported"
_A : str = getattr(snake_case_,"""model.classification.activation.name""","""swish""" )
# config.image_size == getattr(orig_config, 'sampler.bs.crop_size_width', 256)
if is_segmentation_model:
_A : Optional[int] = getattr(snake_case_,"""model.segmentation.output_stride""",16 )
if "_deeplabv3" in task_name:
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_rates""",[12, 24, 36] )
_A : int = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_out_channels""",512 )
_A : str = getattr(snake_case_,"""model.segmentation.deeplabv3.aspp_dropout""",0.1 )
# id2label
_A : List[Any] = """huggingface/label-files"""
_A : List[Any] = json.load(open(hf_hub_download(snake_case_,snake_case_,repo_type="""dataset""" ),"""r""" ) )
_A : str = {int(snake_case_ ): v for k, v in idalabel.items()}
_A : str = idalabel
_A : Dict = {v: k for k, v in idalabel.items()}
return config
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
_A : Any = dct.pop(snake_case_ )
_A : Union[str, Any] = val
def lowerCAmelCase_ ( snake_case_,snake_case_=False ):
if base_model:
_A : Optional[int] = """"""
else:
_A : Dict = """mobilevitv2."""
_A : int = []
for k in state_dict.keys():
if k[:8] == "encoder.":
_A : Any = k[8:]
else:
_A : List[str] = k
if ".block." in k:
_A : Any = k_new.replace(""".block.""",""".""" )
if ".conv." in k:
_A : List[Any] = k_new.replace(""".conv.""",""".convolution.""" )
if ".norm." in k:
_A : Any = k_new.replace(""".norm.""",""".normalization.""" )
if "conv_1." in k:
_A : int = k_new.replace("""conv_1.""",f'''{model_prefix}conv_stem.''' )
for i in [1, 2]:
if f'''layer_{i}.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.''',f'''{model_prefix}encoder.layer.{i-1}.layer.''' )
if ".exp_1x1." in k:
_A : Tuple = k_new.replace(""".exp_1x1.""",""".expand_1x1.""" )
if ".red_1x1." in k:
_A : Optional[int] = k_new.replace(""".red_1x1.""",""".reduce_1x1.""" )
for i in [3, 4, 5]:
if f'''layer_{i}.0.''' in k:
_A : Optional[int] = k_new.replace(f'''layer_{i}.0.''',f'''{model_prefix}encoder.layer.{i-1}.downsampling_layer.''' )
if f'''layer_{i}.1.local_rep.0.''' in k:
_A : Union[str, Any] = k_new.replace(f'''layer_{i}.1.local_rep.0.''',f'''{model_prefix}encoder.layer.{i-1}.conv_kxk.''' )
if f'''layer_{i}.1.local_rep.1.''' in k:
_A : str = k_new.replace(f'''layer_{i}.1.local_rep.1.''',f'''{model_prefix}encoder.layer.{i-1}.conv_1x1.''' )
for i in [3, 4, 5]:
if i == 3:
_A : Optional[int] = [0, 1]
elif i == 4:
_A : Union[str, Any] = [0, 1, 2, 3]
elif i == 5:
_A : Optional[Any] = [0, 1, 2]
for j in j_in:
if f'''layer_{i}.1.global_rep.{j}.''' in k:
_A : Union[str, Any] = k_new.replace(
f'''layer_{i}.1.global_rep.{j}.''',f'''{model_prefix}encoder.layer.{i-1}.transformer.layer.{j}.''' )
if f'''layer_{i}.1.global_rep.{j+1}.''' in k:
_A : List[str] = k_new.replace(
f'''layer_{i}.1.global_rep.{j+1}.''',f'''{model_prefix}encoder.layer.{i-1}.layernorm.''' )
if f'''layer_{i}.1.conv_proj.''' in k:
_A : Optional[Any] = k_new.replace(f'''layer_{i}.1.conv_proj.''',f'''{model_prefix}encoder.layer.{i-1}.conv_projection.''' )
if "pre_norm_attn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_attn.0.""","""layernorm_before.""" )
if "pre_norm_attn.1." in k:
_A : str = k_new.replace("""pre_norm_attn.1.""","""attention.""" )
if "pre_norm_ffn.0." in k:
_A : Optional[Any] = k_new.replace("""pre_norm_ffn.0.""","""layernorm_after.""" )
if "pre_norm_ffn.1." in k:
_A : Dict = k_new.replace("""pre_norm_ffn.1.""","""ffn.conv1.""" )
if "pre_norm_ffn.3." in k:
_A : List[str] = k_new.replace("""pre_norm_ffn.3.""","""ffn.conv2.""" )
if "classifier.1." in k:
_A : List[str] = k_new.replace("""classifier.1.""","""classifier.""" )
if "seg_head." in k:
_A : List[Any] = k_new.replace("""seg_head.""","""segmentation_head.""" )
if ".aspp_layer." in k:
_A : List[Any] = k_new.replace(""".aspp_layer.""",""".""" )
if ".aspp_pool." in k:
_A : Optional[Any] = k_new.replace(""".aspp_pool.""",""".""" )
rename_keys.append((k, k_new) )
return rename_keys
def lowerCAmelCase_ ( snake_case_ ):
_A : Tuple = []
for k in state_dict.keys():
if k.startswith("""seg_head.aux_head.""" ):
keys_to_ignore.append(snake_case_ )
for k in keys_to_ignore:
state_dict.pop(snake_case_,snake_case_ )
def lowerCAmelCase_ ( ):
_A : Dict = """http://images.cocodataset.org/val2017/000000039769.jpg"""
# url = "https://cdn.britannica.com/86/141086-050-9D7C75EE/Gulfstream-G450-business-jet-passengers.jpg"
_A : List[Any] = Image.open(requests.get(snake_case_,stream=snake_case_ ).raw )
return im
@torch.no_grad()
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : List[Any] = get_mobilevitva_config(snake_case_,snake_case_ )
# load original state_dict
_A : Tuple = torch.load(snake_case_,map_location="""cpu""" )
# load huggingface model
if task_name.startswith("""ade20k_""" ) or task_name.startswith("""voc_""" ):
_A : Optional[Any] = MobileViTVaForSemanticSegmentation(snake_case_ ).eval()
_A : str = False
else:
_A : int = MobileViTVaForImageClassification(snake_case_ ).eval()
_A : List[Any] = False
# remove and rename some keys of load the original model
_A : List[Any] = checkpoint
remove_unused_keys(snake_case_ )
_A : Optional[Any] = create_rename_keys(snake_case_,base_model=snake_case_ )
for rename_key_src, rename_key_dest in rename_keys:
rename_key(snake_case_,snake_case_,snake_case_ )
# load modified state_dict
model.load_state_dict(snake_case_ )
# Check outputs on an image, prepared by MobileViTImageProcessor
_A : str = MobileViTImageProcessor(crop_size=config.image_size,size=config.image_size + 32 )
_A : List[Any] = image_processor(images=prepare_img(),return_tensors="""pt""" )
_A : Optional[Any] = model(**snake_case_ )
# verify classification model
if task_name.startswith("""imagenet""" ):
_A : List[Any] = outputs.logits
_A : Optional[int] = logits.argmax(-1 ).item()
print("""Predicted class:""",model.config.idalabel[predicted_class_idx] )
if task_name.startswith("""imagenet1k_256""" ) and config.width_multiplier == 1.0:
# expected_logits for base variant
_A : int = torch.tensor([-1.63_36e00, -7.32_04e-02, -5.18_83e-01] )
assert torch.allclose(logits[0, :3],snake_case_,atol=1e-4 )
Path(snake_case_ ).mkdir(exist_ok=snake_case_ )
print(f'''Saving model {task_name} to {pytorch_dump_folder_path}''' )
model.save_pretrained(snake_case_ )
print(f'''Saving image processor to {pytorch_dump_folder_path}''' )
image_processor.save_pretrained(snake_case_ )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
"--task",
default="imagenet1k_256",
type=str,
help=(
"Name of the task for which the MobileViTV2 model you'd like to convert is trained on . "
"\n Classification (ImageNet-1k)\n - MobileViTV2 (256x256) : imagenet1k_256\n - MobileViTV2 (Trained on 256x256 and Finetuned on 384x384) : imagenet1k_384\n - MobileViTV2 (Trained on ImageNet-21k and Finetuned on ImageNet-1k 256x256) :\n imagenet21k_to_1k_256\n - MobileViTV2 (Trained on ImageNet-21k, Finetuned on ImageNet-1k 256x256, and Finetuned on\n ImageNet-1k 384x384) : imagenet21k_to_1k_384\n Segmentation\n - ADE20K Dataset : ade20k_deeplabv3\n - Pascal VOC 2012 Dataset: voc_deeplabv3\n "
),
choices=[
"imagenet1k_256",
"imagenet1k_384",
"imagenet21k_to_1k_256",
"imagenet21k_to_1k_384",
"ade20k_deeplabv3",
"voc_deeplabv3",
],
)
parser.add_argument(
"--orig_checkpoint_path", required=True, type=str, help="Path to the original state dict (.pt file)."
)
parser.add_argument("--orig_config_path", required=True, type=str, help="Path to the original config file.")
parser.add_argument(
"--pytorch_dump_folder_path", required=True, type=str, help="Path to the output PyTorch model directory."
)
_snake_case = parser.parse_args()
convert_mobilevitva_checkpoint(
args.task, args.orig_checkpoint_path, args.orig_config_path, args.pytorch_dump_folder_path
)
| 26 | 1 |
import random
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
UNetaDConditionModel,
VideoToVideoSDPipeline,
)
from diffusers.utils import floats_tensor, is_xformers_available, skip_mps
from diffusers.utils.testing_utils import enable_full_determinism, slow, torch_device
from ..pipeline_params import (
TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_PARAMS,
)
from ..test_pipelines_common import PipelineTesterMixin
enable_full_determinism()
@skip_mps
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = VideoToVideoSDPipeline
_a = TEXT_GUIDED_IMAGE_VARIATION_PARAMS.union({"video"} ) - {"image", "width", "height"}
_a = TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS.union({"video"} ) - {"image"}
_a = PipelineTesterMixin.required_optional_params - {"latents"}
_a = False
# No `output_type`.
_a = frozenset(
[
"num_inference_steps",
"generator",
"latents",
"return_dict",
"callback",
"callback_steps",
] )
def a__ ( self ) -> Optional[Any]:
torch.manual_seed(0 )
_A : Tuple = UNetaDConditionModel(
block_out_channels=(32, 64, 64, 64) , layers_per_block=2 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""CrossAttnDownBlock3D""", """CrossAttnDownBlock3D""", """CrossAttnDownBlock3D""", """DownBlock3D""") , up_block_types=("""UpBlock3D""", """CrossAttnUpBlock3D""", """CrossAttnUpBlock3D""", """CrossAttnUpBlock3D""") , cross_attention_dim=32 , attention_head_dim=4 , )
_A : int = DDIMScheduler(
beta_start=0.00085 , beta_end=0.012 , beta_schedule="""scaled_linear""" , clip_sample=_a , set_alpha_to_one=_a , )
torch.manual_seed(0 )
_A : str = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
_A : Tuple = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-05 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1000 , hidden_act="""gelu""" , projection_dim=512 , )
_A : Optional[Any] = CLIPTextModel(_a )
_A : Optional[int] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" )
_A : List[Any] = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
}
return components
def a__ ( self , _a , _a=0 ) -> List[Any]:
# 3 frames
_A : List[str] = floats_tensor((1, 3, 3, 32, 32) , rng=random.Random(_a ) ).to(_a )
if str(_a ).startswith("""mps""" ):
_A : Optional[int] = torch.manual_seed(_a )
else:
_A : Any = torch.Generator(device=_a ).manual_seed(_a )
_A : Optional[int] = {
"""prompt""": """A painting of a squirrel eating a burger""",
"""video""": video,
"""generator""": generator,
"""num_inference_steps""": 2,
"""guidance_scale""": 6.0,
"""output_type""": """pt""",
}
return inputs
def a__ ( self ) -> int:
_A : Any = """cpu""" # ensure determinism for the device-dependent torch.Generator
_A : str = self.get_dummy_components()
_A : Optional[int] = VideoToVideoSDPipeline(**_a )
_A : List[Any] = sd_pipe.to(_a )
sd_pipe.set_progress_bar_config(disable=_a )
_A : int = self.get_dummy_inputs(_a )
_A : Optional[int] = """np"""
_A : int = sd_pipe(**_a ).frames
_A : Union[str, Any] = frames[0][-3:, -3:, -1]
assert frames[0].shape == (32, 32, 3)
_A : Dict = np.array([106, 117, 113, 174, 137, 112, 148, 151, 131] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2
@unittest.skipIf(
torch_device != """cuda""" or not is_xformers_available() , reason="""XFormers attention is only available with CUDA and `xformers` installed""" , )
def a__ ( self ) -> Optional[int]:
self._test_xformers_attention_forwardGenerator_pass(test_mean_pixel_difference=_a , expected_max_diff=5e-3 )
@unittest.skip(reason="""Batching needs to be properly figured out first for this pipeline.""" )
def a__ ( self ) -> Tuple:
pass
@unittest.skip(reason="""Batching needs to be properly figured out first for this pipeline.""" )
def a__ ( self ) -> Dict:
pass
@unittest.skip(reason="""`num_images_per_prompt` argument is not supported for this pipeline.""" )
def a__ ( self ) -> List[Any]:
pass
def a__ ( self ) -> List[str]:
return super().test_progress_bar()
@slow
@skip_mps
class lowercase ( unittest.TestCase ):
def a__ ( self ) -> Any:
_A : Union[str, Any] = VideoToVideoSDPipeline.from_pretrained("""cerspense/zeroscope_v2_XL""" , torch_dtype=torch.floataa )
pipe.enable_model_cpu_offload()
# 10 frames
_A : List[Any] = torch.Generator(device="""cpu""" ).manual_seed(0 )
_A : int = torch.randn((1, 10, 3, 1024, 576) , generator=_a )
_A : int = video.to("""cuda""" )
_A : Optional[Any] = """Spiderman is surfing"""
_A : List[Any] = pipe(_a , video=_a , generator=_a , num_inference_steps=3 , output_type="""pt""" ).frames
_A : Dict = np.array([-1.0458984, -1.1279297, -0.9663086, -0.91503906, -0.75097656] )
assert np.abs(video_frames.cpu().numpy()[0, 0, 0, 0, -5:] - expected_array ).sum() < 1e-2
| 26 |
import torch
from diffusers import DPMSolverSDEScheduler
from diffusers.utils import torch_device
from diffusers.utils.testing_utils import require_torchsde
from .test_schedulers import SchedulerCommonTest
@require_torchsde
class lowercase ( UpperCamelCase__ ):
_a = (DPMSolverSDEScheduler,)
_a = 1_0
def a__ ( self , **_a ) -> Optional[Any]:
_A : str = {
"""num_train_timesteps""": 1100,
"""beta_start""": 0.0001,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
"""noise_sampler_seed""": 0,
}
config.update(**_a )
return config
def a__ ( self ) -> Tuple:
for timesteps in [10, 50, 100, 1000]:
self.check_over_configs(num_train_timesteps=_a )
def a__ ( self ) -> Optional[int]:
for beta_start, beta_end in zip([0.00001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ):
self.check_over_configs(beta_start=_a , beta_end=_a )
def a__ ( self ) -> Any:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=_a )
def a__ ( self ) -> Optional[int]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=_a )
def a__ ( self ) -> Optional[int]:
_A : Any = self.scheduler_classes[0]
_A : List[str] = self.get_scheduler_config()
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Dict = self.dummy_model()
_A : Any = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Dict = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : str = model(_a , _a )
_A : List[Any] = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Dict = torch.sum(torch.abs(_a ) )
_A : Dict = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.47821044921875 ) < 1e-2
assert abs(result_mean.item() - 0.2178705964565277 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59352111816406 ) < 1e-2
assert abs(result_mean.item() - 0.22342906892299652 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Optional[Any]:
_A : Dict = self.scheduler_classes[0]
_A : Optional[int] = self.get_scheduler_config(prediction_type="""v_prediction""" )
_A : Optional[Any] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps )
_A : Tuple = self.dummy_model()
_A : int = self.dummy_sample_deter * scheduler.init_noise_sigma
_A : Tuple = sample.to(_a )
for i, t in enumerate(scheduler.timesteps ):
_A : int = scheduler.scale_model_input(_a , _a )
_A : Tuple = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Optional[int] = output.prev_sample
_A : Optional[Any] = torch.sum(torch.abs(_a ) )
_A : List[Any] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 124.77149200439453 ) < 1e-2
assert abs(result_mean.item() - 0.16226289014816284 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 128.1663360595703 ) < 1e-2
assert abs(result_mean.item() - 0.16688326001167297 ) < 1e-3
else:
assert abs(result_sum.item() - 119.8487548828125 ) < 1e-2
assert abs(result_mean.item() - 0.1560530662536621 ) < 1e-3
def a__ ( self ) -> List[str]:
_A : Union[str, Any] = self.scheduler_classes[0]
_A : List[Any] = self.get_scheduler_config()
_A : List[str] = scheduler_class(**_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Union[str, Any] = self.dummy_model()
_A : Optional[Any] = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
_A : int = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : Dict = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : str = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 167.46957397460938 ) < 1e-2
assert abs(result_mean.item() - 0.21805934607982635 ) < 1e-3
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 171.59353637695312 ) < 1e-2
assert abs(result_mean.item() - 0.22342908382415771 ) < 1e-3
else:
assert abs(result_sum.item() - 162.52383422851562 ) < 1e-2
assert abs(result_mean.item() - 0.211619570851326 ) < 1e-3
def a__ ( self ) -> Union[str, Any]:
_A : List[Any] = self.scheduler_classes[0]
_A : Optional[Any] = self.get_scheduler_config()
_A : int = scheduler_class(**_a , use_karras_sigmas=_a )
scheduler.set_timesteps(self.num_inference_steps , device=_a )
_A : Optional[Any] = self.dummy_model()
_A : Dict = self.dummy_sample_deter.to(_a ) * scheduler.init_noise_sigma
_A : str = sample.to(_a )
for t in scheduler.timesteps:
_A : Optional[int] = scheduler.scale_model_input(_a , _a )
_A : List[Any] = model(_a , _a )
_A : Dict = scheduler.step(_a , _a , _a )
_A : List[str] = output.prev_sample
_A : str = torch.sum(torch.abs(_a ) )
_A : List[str] = torch.mean(torch.abs(_a ) )
if torch_device in ["mps"]:
assert abs(result_sum.item() - 176.66974135742188 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
elif torch_device in ["cuda"]:
assert abs(result_sum.item() - 177.63653564453125 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
else:
assert abs(result_sum.item() - 170.3135223388672 ) < 1e-2
assert abs(result_mean.item() - 0.23003872730981811 ) < 1e-2
| 26 | 1 |
import argparse
import logging
from collections import namedtuple
import torch
from model_bertabs import BertAbsSummarizer
from models.model_builder import AbsSummarizer # The authors' implementation
from transformers import BertTokenizer
logging.basicConfig(level=logging.INFO)
_snake_case = logging.getLogger(__name__)
_snake_case = "Hello world! cécé herlolip"
_snake_case = namedtuple(
"BertAbsConfig",
[
"temp_dir",
"large",
"use_bert_emb",
"finetune_bert",
"encoder",
"share_emb",
"max_pos",
"enc_layers",
"enc_hidden_size",
"enc_heads",
"enc_ff_size",
"enc_dropout",
"dec_layers",
"dec_hidden_size",
"dec_heads",
"dec_ff_size",
"dec_dropout",
],
)
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_A : Optional[Any] = BertAbsConfig(
temp_dir=""".""",finetune_bert=snake_case_,large=snake_case_,share_emb=snake_case_,use_bert_emb=snake_case_,encoder="""bert""",max_pos=512,enc_layers=6,enc_hidden_size=512,enc_heads=8,enc_ff_size=512,enc_dropout=0.2,dec_layers=6,dec_hidden_size=768,dec_heads=8,dec_ff_size=2048,dec_dropout=0.2,)
_A : Optional[int] = torch.load(snake_case_,lambda snake_case_,snake_case_ : storage )
_A : Optional[int] = AbsSummarizer(snake_case_,torch.device("""cpu""" ),snake_case_ )
original.eval()
_A : List[str] = BertAbsSummarizer(snake_case_,torch.device("""cpu""" ) )
new_model.eval()
# -------------------
# Convert the weights
# -------------------
logging.info("""convert the model""" )
new_model.bert.load_state_dict(original.bert.state_dict() )
new_model.decoder.load_state_dict(original.decoder.state_dict() )
new_model.generator.load_state_dict(original.generator.state_dict() )
# ----------------------------------
# Make sure the outpus are identical
# ----------------------------------
logging.info("""Make sure that the models' outputs are identical""" )
_A : List[str] = BertTokenizer.from_pretrained("""bert-base-uncased""" )
# prepare the model inputs
_A : List[Any] = tokenizer.encode("""This is sample éàalj'-.""" )
encoder_input_ids.extend([tokenizer.pad_token_id] * (512 - len(snake_case_ )) )
_A : int = torch.tensor(snake_case_ ).unsqueeze(0 )
_A : Any = tokenizer.encode("""This is sample 3 éàalj'-.""" )
decoder_input_ids.extend([tokenizer.pad_token_id] * (512 - len(snake_case_ )) )
_A : List[str] = torch.tensor(snake_case_ ).unsqueeze(0 )
# failsafe to make sure the weights reset does not affect the
# loaded weights.
assert torch.max(torch.abs(original.generator[0].weight - new_model.generator[0].weight ) ) == 0
# forward pass
_A : List[str] = encoder_input_ids
_A : int = decoder_input_ids
_A : Tuple = None
_A : List[Any] = None
_A : Tuple = None
_A : Tuple = None
_A : List[str] = None
# The original model does not apply the geneator layer immediatly but rather in
# the beam search (where it combines softmax + linear layer). Since we already
# apply the softmax in our generation process we only apply the linear layer here.
# We make sure that the outputs of the full stack are identical
_A : List[str] = original(snake_case_,snake_case_,snake_case_,snake_case_,snake_case_,snake_case_,snake_case_ )[0]
_A : List[str] = original.generator(snake_case_ )
_A : str = new_model(
snake_case_,snake_case_,snake_case_,snake_case_,snake_case_ )[0]
_A : Optional[Any] = new_model.generator(snake_case_ )
_A : int = torch.max(torch.abs(output_converted_model - output_original_model ) ).item()
print("""Maximum absolute difference beween weights: {:.2f}""".format(snake_case_ ) )
_A : str = torch.max(torch.abs(output_converted_generator - output_original_generator ) ).item()
print("""Maximum absolute difference beween weights: {:.2f}""".format(snake_case_ ) )
_A : Optional[Any] = torch.allclose(snake_case_,snake_case_,atol=1e-3 )
if are_identical:
logging.info("""all weights are equal up to 1e-3""" )
else:
raise ValueError("""the weights are different. The new model is likely different from the original one.""" )
# The model has been saved with torch.save(model) and this is bound to the exact
# directory structure. We save the state_dict instead.
logging.info("""saving the model's state dictionary""" )
torch.save(
new_model.state_dict(),"""./bertabs-finetuned-cnndm-extractive-abstractive-summarization/pytorch_model.bin""" )
if __name__ == "__main__":
_snake_case = argparse.ArgumentParser()
parser.add_argument(
"--bertabs_checkpoint_path",
default=None,
type=str,
required=True,
help="Path the official PyTorch dump.",
)
parser.add_argument(
"--pytorch_dump_folder_path",
default=None,
type=str,
required=True,
help="Path to the output PyTorch model.",
)
_snake_case = parser.parse_args()
convert_bertabs_checkpoints(
args.bertabs_checkpoint_path,
args.pytorch_dump_folder_path,
)
| 26 |
# DISCLAIMER: This file is strongly influenced by https://github.com/yang-song/score_sde_pytorch
import math
from typing import Union
import torch
from ..configuration_utils import ConfigMixin, register_to_config
from ..utils import randn_tensor
from .scheduling_utils import SchedulerMixin
class lowercase ( UpperCamelCase__,UpperCamelCase__ ):
_a = 1
@register_to_config
def __init__( self , _a=2000 , _a=0.1 , _a=20 , _a=1e-3 ) -> List[Any]:
_A : Dict = None
_A : List[Any] = None
_A : Dict = None
def a__ ( self , _a , _a = None ) -> Union[str, Any]:
_A : Union[str, Any] = torch.linspace(1 , self.config.sampling_eps , _a , device=_a )
def a__ ( self , _a , _a , _a , _a=None ) -> Dict:
if self.timesteps is None:
raise ValueError(
"""`self.timesteps` is not set, you need to run 'set_timesteps' after creating the scheduler""" )
# TODO(Patrick) better comments + non-PyTorch
# postprocess model score
_A : Any = (
-0.25 * t**2 * (self.config.beta_max - self.config.beta_min) - 0.5 * t * self.config.beta_min
)
_A : List[Any] = torch.sqrt(1.0 - torch.exp(2.0 * log_mean_coeff ) )
_A : List[str] = std.flatten()
while len(std.shape ) < len(score.shape ):
_A : List[Any] = std.unsqueeze(-1 )
_A : int = -score / std
# compute
_A : Tuple = -1.0 / len(self.timesteps )
_A : str = self.config.beta_min + t * (self.config.beta_max - self.config.beta_min)
_A : List[str] = beta_t.flatten()
while len(beta_t.shape ) < len(x.shape ):
_A : Union[str, Any] = beta_t.unsqueeze(-1 )
_A : Tuple = -0.5 * beta_t * x
_A : Tuple = torch.sqrt(_a )
_A : Dict = drift - diffusion**2 * score
_A : Dict = x + drift * dt
# add noise
_A : Any = randn_tensor(x.shape , layout=x.layout , generator=_a , device=x.device , dtype=x.dtype )
_A : str = x_mean + diffusion * math.sqrt(-dt ) * noise
return x, x_mean
def __len__( self ) -> Optional[Any]:
return self.config.num_train_timesteps
| 26 | 1 |
# Copyright 2023 The HuggingFace Team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available
_snake_case = {
"configuration_xmod": [
"XMOD_PRETRAINED_CONFIG_ARCHIVE_MAP",
"XmodConfig",
"XmodOnnxConfig",
],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
_snake_case = [
"XMOD_PRETRAINED_MODEL_ARCHIVE_LIST",
"XmodForCausalLM",
"XmodForMaskedLM",
"XmodForMultipleChoice",
"XmodForQuestionAnswering",
"XmodForSequenceClassification",
"XmodForTokenClassification",
"XmodModel",
"XmodPreTrainedModel",
]
if TYPE_CHECKING:
from .configuration_xmod import XMOD_PRETRAINED_CONFIG_ARCHIVE_MAP, XmodConfig, XmodOnnxConfig
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_xmod import (
XMOD_PRETRAINED_MODEL_ARCHIVE_LIST,
XmodForCausalLM,
XmodForMaskedLM,
XmodForMultipleChoice,
XmodForQuestionAnswering,
XmodForSequenceClassification,
XmodForTokenClassification,
XmodModel,
XmodPreTrainedModel,
)
else:
import sys
_snake_case = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
| 26 |
import os
from shutil import copyfile
from typing import List, Optional, Tuple
from ...tokenization_utils import AddedToken
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import is_sentencepiece_available, logging
if is_sentencepiece_available():
from .tokenization_fnet import FNetTokenizer
else:
_snake_case = None
_snake_case = logging.get_logger(__name__)
_snake_case = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.json"}
_snake_case = {
"vocab_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/spiece.model",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/spiece.model",
},
"tokenizer_file": {
"google/fnet-base": "https://huggingface.co/google/fnet-base/resolve/main/tokenizer.json",
"google/fnet-large": "https://huggingface.co/google/fnet-large/resolve/main/tokenizer.json",
},
}
_snake_case = {
"google/fnet-base": 512,
"google/fnet-large": 512,
}
_snake_case = "▁"
class lowercase ( UpperCamelCase__ ):
_a = VOCAB_FILES_NAMES
_a = PRETRAINED_VOCAB_FILES_MAP
_a = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
_a = ["input_ids", "token_type_ids"]
_a = FNetTokenizer
def __init__( self , _a=None , _a=None , _a=False , _a=True , _a=True , _a="<unk>" , _a="[SEP]" , _a="<pad>" , _a="[CLS]" , _a="[MASK]" , **_a , ) -> Optional[int]:
# Mask token behave like a normal word, i.e. include the space before it and
# is included in the raw text, there should be a match in a non-normalized sentence.
_A : int = (
AddedToken(_a , lstrip=_a , rstrip=_a , normalized=_a )
if isinstance(_a , _a )
else mask_token
)
super().__init__(
_a , tokenizer_file=_a , do_lower_case=_a , remove_space=_a , keep_accents=_a , unk_token=_a , sep_token=_a , pad_token=_a , cls_token=_a , mask_token=_a , **_a , )
_A : Optional[int] = do_lower_case
_A : List[Any] = remove_space
_A : str = keep_accents
_A : int = vocab_file
_A : int = False if not self.vocab_file else True
def a__ ( self , _a , _a = None ) -> List[int]:
_A : str = [self.sep_token_id]
_A : Dict = [self.cls_token_id]
if token_ids_a is None:
return cls + token_ids_a + sep
return cls + token_ids_a + sep + token_ids_a + sep
def a__ ( self , _a , _a = None ) -> List[int]:
_A : Any = [self.sep_token_id]
_A : Any = [self.cls_token_id]
if token_ids_a is None:
return len(cls + token_ids_a + sep ) * [0]
return len(cls + token_ids_a + sep ) * [0] + len(token_ids_a + sep ) * [1]
def a__ ( self , _a , _a = None ) -> Tuple[str]:
if not os.path.isdir(_a ):
logger.error(F'''Vocabulary path ({save_directory}) should be a directory''' )
return
_A : List[str] = os.path.join(
_a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
if os.path.abspath(self.vocab_file ) != os.path.abspath(_a ):
copyfile(self.vocab_file , _a )
return (out_vocab_file,)
| 26 | 1 |
import unittest
from typing import Dict, List, Optional, Union
import numpy as np
from transformers.testing_utils import require_torch, require_vision
from transformers.utils import is_torch_available, is_vision_available
from ...test_image_processing_common import ImageProcessingSavingTestMixin, prepare_image_inputs
if is_torch_available():
import torch
if is_vision_available():
from PIL import Image
from transformers import BridgeTowerImageProcessor
class lowercase ( unittest.TestCase ):
def __init__( self , _a , _a = True , _a = None , _a = 32 , _a = True , _a = 1 / 255 , _a = True , _a = True , _a = [0.48145466, 0.4578275, 0.40821073] , _a = [0.26862954, 0.26130258, 0.27577711] , _a = True , _a=7 , _a=30 , _a=400 , _a=3 , ) -> List[Any]:
_A : List[str] = parent
_A : Any = do_resize
_A : Tuple = size if size is not None else {"""shortest_edge""": 288}
_A : Dict = size_divisor
_A : Optional[Any] = do_rescale
_A : int = rescale_factor
_A : Dict = do_normalize
_A : Union[str, Any] = do_center_crop
_A : Any = image_mean
_A : int = image_std
_A : Optional[Any] = do_pad
_A : List[str] = batch_size
_A : Tuple = num_channels
_A : Optional[int] = min_resolution
_A : int = max_resolution
def a__ ( self ) -> List[str]:
return {
"image_mean": self.image_mean,
"image_std": self.image_std,
"do_normalize": self.do_normalize,
"do_resize": self.do_resize,
"size": self.size,
"size_divisor": self.size_divisor,
}
def a__ ( self , _a , _a=False ) -> Optional[int]:
if not batched:
_A : Optional[int] = self.size["""shortest_edge"""]
_A : Tuple = image_inputs[0]
if isinstance(_a , Image.Image ):
_A , _A : Dict = image.size
else:
_A , _A : Union[str, Any] = image.shape[1], image.shape[2]
_A : int = size / min(_a , _a )
if h < w:
_A , _A : int = size, scale * w
else:
_A , _A : Dict = scale * h, size
_A : int = int((1333 / 800) * size )
if max(_a , _a ) > max_size:
_A : Union[str, Any] = max_size / max(_a , _a )
_A : Any = newh * scale
_A : Union[str, Any] = neww * scale
_A , _A : str = int(newh + 0.5 ), int(neww + 0.5 )
_A , _A : Any = (
newh // self.size_divisor * self.size_divisor,
neww // self.size_divisor * self.size_divisor,
)
else:
_A : Dict = []
for image in image_inputs:
_A , _A : Dict = self.get_expected_values([image] )
expected_values.append((expected_height, expected_width) )
_A : List[Any] = max(_a , key=lambda _a : item[0] )[0]
_A : Dict = max(_a , key=lambda _a : item[1] )[1]
return expected_height, expected_width
@require_torch
@require_vision
class lowercase ( UpperCamelCase__,unittest.TestCase ):
_a = BridgeTowerImageProcessor if is_vision_available() else None
def a__ ( self ) -> Any:
_A : int = BridgeTowerImageProcessingTester(self )
@property
def a__ ( self ) -> List[Any]:
return self.image_processor_tester.prepare_image_processor_dict()
def a__ ( self ) -> List[str]:
_A : int = self.image_processing_class(**self.image_processor_dict )
self.assertTrue(hasattr(_a , """image_mean""" ) )
self.assertTrue(hasattr(_a , """image_std""" ) )
self.assertTrue(hasattr(_a , """do_normalize""" ) )
self.assertTrue(hasattr(_a , """do_resize""" ) )
self.assertTrue(hasattr(_a , """size""" ) )
self.assertTrue(hasattr(_a , """size_divisor""" ) )
def a__ ( self ) -> List[Any]:
pass
def a__ ( self ) -> Tuple:
# Initialize image processor
_A : str = self.image_processing_class(**self.image_processor_dict )
# create random PIL images
_A : Tuple = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a )
for image in image_inputs:
self.assertIsInstance(_a , Image.Image )
# Test not batched input
_A : Union[str, Any] = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
_A , _A : Optional[Any] = self.image_processor_tester.get_expected_values(_a )
self.assertEqual(
encoded_images.shape , (1, self.image_processor_tester.num_channels, expected_height, expected_width) , )
# Test batched
_A : Tuple = image_processing(_a , return_tensors="""pt""" ).pixel_values
_A , _A : int = self.image_processor_tester.get_expected_values(_a , batched=_a )
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
expected_height,
expected_width,
) , )
def a__ ( self ) -> Tuple:
# Initialize image processor
_A : Dict = self.image_processing_class(**self.image_processor_dict )
# create random numpy tensors
_A : str = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , numpify=_a )
for image in image_inputs:
self.assertIsInstance(_a , np.ndarray )
# Test not batched input
_A : Union[str, Any] = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
_A , _A : Dict = self.image_processor_tester.get_expected_values(_a )
self.assertEqual(
encoded_images.shape , (1, self.image_processor_tester.num_channels, expected_height, expected_width) , )
# Test batched
_A : List[Any] = image_processing(_a , return_tensors="""pt""" ).pixel_values
_A , _A : Tuple = self.image_processor_tester.get_expected_values(_a , batched=_a )
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
expected_height,
expected_width,
) , )
def a__ ( self ) -> Any:
# Initialize image processor
_A : Optional[Any] = self.image_processing_class(**self.image_processor_dict )
# create random PyTorch tensors
_A : Any = prepare_image_inputs(self.image_processor_tester , equal_resolution=_a , torchify=_a )
for image in image_inputs:
self.assertIsInstance(_a , torch.Tensor )
# Test not batched input
_A : Optional[int] = image_processing(image_inputs[0] , return_tensors="""pt""" ).pixel_values
_A , _A : Any = self.image_processor_tester.get_expected_values(_a )
self.assertEqual(
encoded_images.shape , (1, self.image_processor_tester.num_channels, expected_height, expected_width) , )
# Test batched
_A : Optional[Any] = image_processing(_a , return_tensors="""pt""" ).pixel_values
_A , _A : str = self.image_processor_tester.get_expected_values(_a , batched=_a )
self.assertEqual(
encoded_images.shape , (
self.image_processor_tester.batch_size,
self.image_processor_tester.num_channels,
expected_height,
expected_width,
) , )
| 26 |
from math import asin, atan, cos, radians, sin, sqrt, tan
_snake_case = 6_3_7_8_1_3_7.0
_snake_case = 6_3_5_6_7_5_2.3_1_4_2_4_5
_snake_case = 6378137
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_,snake_case_ ):
_A : Any = (AXIS_A - AXIS_B) / AXIS_A
_A : Optional[int] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : List[str] = atan((1 - flattening) * tan(radians(snake_case_ ) ) )
_A : Optional[Any] = radians(snake_case_ )
_A : str = radians(snake_case_ )
# Equation
_A : Dict = sin((phi_a - phi_a) / 2 )
_A : List[str] = sin((lambda_a - lambda_a) / 2 )
# Square both values
sin_sq_phi *= sin_sq_phi
sin_sq_lambda *= sin_sq_lambda
_A : Optional[int] = sqrt(sin_sq_phi + (cos(snake_case_ ) * cos(snake_case_ ) * sin_sq_lambda) )
return 2 * RADIUS * asin(snake_case_ )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 26 | 1 |
import argparse
import os
import evaluate
import torch
from datasets import load_dataset
from torch.optim import AdamW
from torch.utils.data import DataLoader
from transformers import AutoModelForSequenceClassification, AutoTokenizer, get_linear_schedule_with_warmup, set_seed
from accelerate import Accelerator, DistributedType
########################################################################
# This is a fully working simple example to use Accelerate
# and perform gradient accumulation
#
# This example trains a Bert base model on GLUE MRPC
# in any of the following settings (with the same script):
# - single CPU or single GPU
# - multi GPUS (using PyTorch distributed mode)
# - (multi) TPUs
# - fp16 (mixed-precision) or fp32 (normal precision)
#
# To run it in each of these various modes, follow the instructions
# in the readme for examples:
# https://github.com/huggingface/accelerate/tree/main/examples
#
########################################################################
_snake_case = 16
_snake_case = 32
def lowerCAmelCase_ ( snake_case_,snake_case_ = 16 ):
_A : List[Any] = AutoTokenizer.from_pretrained("""bert-base-cased""" )
_A : Any = load_dataset("""glue""","""mrpc""" )
def tokenize_function(snake_case_ ):
# max_length=None => use the model max length (it's actually the default)
_A : int = tokenizer(examples["""sentence1"""],examples["""sentence2"""],truncation=snake_case_,max_length=snake_case_ )
return outputs
# Apply the method we just defined to all the examples in all the splits of the dataset
# starting with the main process first:
with accelerator.main_process_first():
_A : Dict = datasets.map(
snake_case_,batched=snake_case_,remove_columns=["""idx""", """sentence1""", """sentence2"""],)
# We also rename the 'label' column to 'labels' which is the expected name for labels by the models of the
# transformers library
_A : List[Any] = tokenized_datasets.rename_column("""label""","""labels""" )
def collate_fn(snake_case_ ):
# On TPU it's best to pad everything to the same length or training will be very slow.
_A : Optional[Any] = 128 if accelerator.distributed_type == DistributedType.TPU else None
# When using mixed precision we want round multiples of 8/16
if accelerator.mixed_precision == "fp8":
_A : Any = 16
elif accelerator.mixed_precision != "no":
_A : int = 8
else:
_A : str = None
return tokenizer.pad(
snake_case_,padding="""longest""",max_length=snake_case_,pad_to_multiple_of=snake_case_,return_tensors="""pt""",)
# Instantiate dataloaders.
_A : Dict = DataLoader(
tokenized_datasets["""train"""],shuffle=snake_case_,collate_fn=snake_case_,batch_size=snake_case_ )
_A : str = DataLoader(
tokenized_datasets["""validation"""],shuffle=snake_case_,collate_fn=snake_case_,batch_size=snake_case_ )
return train_dataloader, eval_dataloader
# For testing only
if os.environ.get("TESTING_MOCKED_DATALOADERS", None) == "1":
from accelerate.test_utils.training import mocked_dataloaders
_snake_case = mocked_dataloaders # noqa: F811
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
# For testing only
if os.environ.get("""TESTING_MOCKED_DATALOADERS""",snake_case_ ) == "1":
_A : Optional[Any] = 2
# New Code #
_A : str = int(args.gradient_accumulation_steps )
# Initialize accelerator
_A : Dict = Accelerator(
cpu=args.cpu,mixed_precision=args.mixed_precision,gradient_accumulation_steps=snake_case_ )
if accelerator.distributed_type == DistributedType.TPU and gradient_accumulation_steps > 1:
raise NotImplementedError(
"""Gradient accumulation on TPUs is currently not supported. Pass `gradient_accumulation_steps=1`""" )
# Sample hyper-parameters for learning rate, batch size, seed and a few other HPs
_A : int = config["""lr"""]
_A : str = int(config["""num_epochs"""] )
_A : Tuple = int(config["""seed"""] )
_A : Optional[Any] = int(config["""batch_size"""] )
_A : int = evaluate.load("""glue""","""mrpc""" )
set_seed(snake_case_ )
_A , _A : Dict = get_dataloaders(snake_case_,snake_case_ )
# Instantiate the model (we build the model here so that the seed also control new weights initialization)
_A : Dict = AutoModelForSequenceClassification.from_pretrained("""bert-base-cased""",return_dict=snake_case_ )
# We could avoid this line since the accelerator is set with `device_placement=True` (default value).
# Note that if you are placing tensors on devices manually, this line absolutely needs to be before the optimizer
# creation otherwise training will not work on TPU (`accelerate` will kindly throw an error to make us aware of that).
_A : List[Any] = model.to(accelerator.device )
# Instantiate optimizer
_A : Any = AdamW(params=model.parameters(),lr=snake_case_ )
# Instantiate scheduler
_A : Optional[Any] = get_linear_schedule_with_warmup(
optimizer=snake_case_,num_warmup_steps=100,num_training_steps=(len(snake_case_ ) * num_epochs),)
# Prepare everything
# There is no specific order to remember, we just need to unpack the objects in the same order we gave them to the
# prepare method.
_A , _A , _A , _A , _A : List[Any] = accelerator.prepare(
snake_case_,snake_case_,snake_case_,snake_case_,snake_case_ )
# Now we train the model
for epoch in range(snake_case_ ):
model.train()
for step, batch in enumerate(snake_case_ ):
# We could avoid this line since we set the accelerator with `device_placement=True`.
batch.to(accelerator.device )
# New code #
# We use the new `accumulate` context manager to perform gradient accumulation
# We also currently do not support TPUs nor advise it as bugs were found on the XLA side when running our tests.
with accelerator.accumulate(snake_case_ ):
_A : List[Any] = model(**snake_case_ )
_A : Dict = output.loss
accelerator.backward(snake_case_ )
optimizer.step()
lr_scheduler.step()
optimizer.zero_grad()
model.eval()
for step, batch in enumerate(snake_case_ ):
# We could avoid this line since we set the accelerator with `device_placement=True`.
batch.to(accelerator.device )
with torch.no_grad():
_A : Dict = model(**snake_case_ )
_A : Any = outputs.logits.argmax(dim=-1 )
_A , _A : Tuple = accelerator.gather_for_metrics((predictions, batch["""labels"""]) )
metric.add_batch(
predictions=snake_case_,references=snake_case_,)
_A : int = metric.compute()
# Use accelerator.print to print only on the main process.
accelerator.print(f'''epoch {epoch}:''',snake_case_ )
def lowerCAmelCase_ ( ):
_A : Any = argparse.ArgumentParser(description="""Simple example of training script.""" )
parser.add_argument(
"""--mixed_precision""",type=snake_case_,default=snake_case_,choices=["""no""", """fp16""", """bf16""", """fp8"""],help="""Whether to use mixed precision. Choose"""
"""between fp16 and bf16 (bfloat16). Bf16 requires PyTorch >= 1.10."""
"""and an Nvidia Ampere GPU.""",)
# New Code #
parser.add_argument(
"""--gradient_accumulation_steps""",type=snake_case_,default=1,help="""The number of minibatches to be ran before gradients are accumulated.""",)
parser.add_argument("""--cpu""",action="""store_true""",help="""If passed, will train on the CPU.""" )
_A : Tuple = parser.parse_args()
_A : List[str] = {"""lr""": 2e-5, """num_epochs""": 3, """seed""": 42, """batch_size""": 16}
training_function(snake_case_,snake_case_ )
if __name__ == "__main__":
main()
| 26 |
from __future__ import absolute_import, division, print_function, unicode_literals
from torch import nn
from torch.nn import CrossEntropyLoss, MSELoss
from transformers import RobertaConfig
from transformers.file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
from transformers.models.roberta.modeling_roberta import (
ROBERTA_INPUTS_DOCSTRING,
ROBERTA_START_DOCSTRING,
RobertaEmbeddings,
)
from .modeling_highway_bert import BertPreTrainedModel, DeeBertModel, HighwayException, entropy
@add_start_docstrings(
"The RoBERTa Model transformer with early exiting (DeeRoBERTa). ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> Optional[int]:
super().__init__(_a )
_A : Union[str, Any] = RobertaEmbeddings(_a )
self.init_weights()
@add_start_docstrings(
"RoBERTa Model (with early exiting - DeeRoBERTa) with a classifier on top,\n also takes care of multi-layer training. ",UpperCamelCase__,)
class lowercase ( UpperCamelCase__ ):
_a = RobertaConfig
_a = "roberta"
def __init__( self , _a ) -> str:
super().__init__(_a )
_A : Any = config.num_labels
_A : Dict = config.num_hidden_layers
_A : List[str] = DeeRobertaModel(_a )
_A : int = nn.Dropout(config.hidden_dropout_prob )
_A : int = nn.Linear(config.hidden_size , self.config.num_labels )
@add_start_docstrings_to_model_forward(_a )
def a__ ( self , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=None , _a=-1 , _a=False , ) -> Any:
_A : Optional[int] = self.num_layers
try:
_A : List[str] = self.roberta(
_a , attention_mask=_a , token_type_ids=_a , position_ids=_a , head_mask=_a , inputs_embeds=_a , )
_A : List[str] = outputs[1]
_A : List[str] = self.dropout(_a )
_A : Optional[Any] = self.classifier(_a )
_A : List[Any] = (logits,) + outputs[2:] # add hidden states and attention if they are here
except HighwayException as e:
_A : List[Any] = e.message
_A : Optional[int] = e.exit_layer
_A : Optional[int] = outputs[0]
if not self.training:
_A : int = entropy(_a )
_A : int = []
_A : int = []
if labels is not None:
if self.num_labels == 1:
# We are doing regression
_A : Union[str, Any] = MSELoss()
_A : Tuple = loss_fct(logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Dict = loss_fct(logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
# work with highway exits
_A : Optional[Any] = []
for highway_exit in outputs[-1]:
_A : Tuple = highway_exit[0]
if not self.training:
highway_logits_all.append(_a )
highway_entropy.append(highway_exit[2] )
if self.num_labels == 1:
# We are doing regression
_A : List[str] = MSELoss()
_A : Optional[int] = loss_fct(highway_logits.view(-1 ) , labels.view(-1 ) )
else:
_A : List[Any] = CrossEntropyLoss()
_A : Tuple = loss_fct(highway_logits.view(-1 , self.num_labels ) , labels.view(-1 ) )
highway_losses.append(_a )
if train_highway:
_A : Dict = (sum(highway_losses[:-1] ),) + outputs
# exclude the final highway, of course
else:
_A : int = (loss,) + outputs
if not self.training:
_A : Optional[Any] = outputs + ((original_entropy, highway_entropy), exit_layer)
if output_layer >= 0:
_A : Union[str, Any] = (
(outputs[0],) + (highway_logits_all[output_layer],) + outputs[2:]
) # use the highway of the last layer
return outputs # (loss), logits, (hidden_states), (attentions), entropy
| 26 | 1 |
import os
import re
import sys
import traceback
import warnings
from pathlib import Path
from typing import Dict, Optional, Union
from uuid import uuida
from huggingface_hub import HfFolder, ModelCard, ModelCardData, hf_hub_download, whoami
from huggingface_hub.file_download import REGEX_COMMIT_HASH
from huggingface_hub.utils import (
EntryNotFoundError,
RepositoryNotFoundError,
RevisionNotFoundError,
is_jinja_available,
)
from packaging import version
from requests import HTTPError
from .. import __version__
from .constants import (
DEPRECATED_REVISION_ARGS,
DIFFUSERS_CACHE,
HUGGINGFACE_CO_RESOLVE_ENDPOINT,
SAFETENSORS_WEIGHTS_NAME,
WEIGHTS_NAME,
)
from .import_utils import (
ENV_VARS_TRUE_VALUES,
_flax_version,
_jax_version,
_onnxruntime_version,
_torch_version,
is_flax_available,
is_onnx_available,
is_torch_available,
)
from .logging import get_logger
_snake_case = get_logger(__name__)
_snake_case = Path(__file__).parent / "model_card_template.md"
_snake_case = uuida().hex
_snake_case = os.getenv("HF_HUB_OFFLINE", "").upper() in ENV_VARS_TRUE_VALUES
_snake_case = os.getenv("DISABLE_TELEMETRY", "").upper() in ENV_VARS_TRUE_VALUES
_snake_case = HUGGINGFACE_CO_RESOLVE_ENDPOINT + "/api/telemetry/"
def lowerCAmelCase_ ( snake_case_ = None ):
_A : List[Any] = f'''diffusers/{__version__}; python/{sys.version.split()[0]}; session_id/{SESSION_ID}'''
if DISABLE_TELEMETRY or HF_HUB_OFFLINE:
return ua + "; telemetry/off"
if is_torch_available():
ua += f'''; torch/{_torch_version}'''
if is_flax_available():
ua += f'''; jax/{_jax_version}'''
ua += f'''; flax/{_flax_version}'''
if is_onnx_available():
ua += f'''; onnxruntime/{_onnxruntime_version}'''
# CI will set this value to True
if os.environ.get("""DIFFUSERS_IS_CI""","""""" ).upper() in ENV_VARS_TRUE_VALUES:
ua += "; is_ci/true"
if isinstance(snake_case_,snake_case_ ):
ua += "; " + "; ".join(f'''{k}/{v}''' for k, v in user_agent.items() )
elif isinstance(snake_case_,snake_case_ ):
ua += "; " + user_agent
return ua
def lowerCAmelCase_ ( snake_case_,snake_case_ = None,snake_case_ = None ):
if token is None:
_A : Union[str, Any] = HfFolder.get_token()
if organization is None:
_A : Any = whoami(snake_case_ )["""name"""]
return f'''{username}/{model_id}'''
else:
return f'''{organization}/{model_id}'''
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if not is_jinja_available():
raise ValueError(
"""Modelcard rendering is based on Jinja templates."""
""" Please make sure to have `jinja` installed before using `create_model_card`."""
""" To install it, please run `pip install Jinja2`.""" )
if hasattr(snake_case_,"""local_rank""" ) and args.local_rank not in [-1, 0]:
return
_A : Any = args.hub_token if hasattr(snake_case_,"""hub_token""" ) else None
_A : int = get_full_repo_name(snake_case_,token=snake_case_ )
_A : Optional[Any] = ModelCard.from_template(
card_data=ModelCardData( # Card metadata object that will be converted to YAML block
language="""en""",license="""apache-2.0""",library_name="""diffusers""",tags=[],datasets=args.dataset_name,metrics=[],),template_path=snake_case_,model_name=snake_case_,repo_name=snake_case_,dataset_name=args.dataset_name if hasattr(snake_case_,"""dataset_name""" ) else None,learning_rate=args.learning_rate,train_batch_size=args.train_batch_size,eval_batch_size=args.eval_batch_size,gradient_accumulation_steps=(
args.gradient_accumulation_steps if hasattr(snake_case_,"""gradient_accumulation_steps""" ) else None
),adam_betaa=args.adam_betaa if hasattr(snake_case_,"""adam_beta1""" ) else None,adam_betaa=args.adam_betaa if hasattr(snake_case_,"""adam_beta2""" ) else None,adam_weight_decay=args.adam_weight_decay if hasattr(snake_case_,"""adam_weight_decay""" ) else None,adam_epsilon=args.adam_epsilon if hasattr(snake_case_,"""adam_epsilon""" ) else None,lr_scheduler=args.lr_scheduler if hasattr(snake_case_,"""lr_scheduler""" ) else None,lr_warmup_steps=args.lr_warmup_steps if hasattr(snake_case_,"""lr_warmup_steps""" ) else None,ema_inv_gamma=args.ema_inv_gamma if hasattr(snake_case_,"""ema_inv_gamma""" ) else None,ema_power=args.ema_power if hasattr(snake_case_,"""ema_power""" ) else None,ema_max_decay=args.ema_max_decay if hasattr(snake_case_,"""ema_max_decay""" ) else None,mixed_precision=args.mixed_precision,)
_A : int = os.path.join(args.output_dir,"""README.md""" )
model_card.save(snake_case_ )
def lowerCAmelCase_ ( snake_case_,snake_case_ = None ):
if resolved_file is None or commit_hash is not None:
return commit_hash
_A : int = str(Path(snake_case_ ).as_posix() )
_A : List[Any] = re.search(r"""snapshots/([^/]+)/""",snake_case_ )
if search is None:
return None
_A : int = search.groups()[0]
return commit_hash if REGEX_COMMIT_HASH.match(snake_case_ ) else None
# Old default cache path, potentially to be migrated.
# This logic was more or less taken from `transformers`, with the following differences:
# - Diffusers doesn't use custom environment variables to specify the cache path.
# - There is no need to migrate the cache format, just move the files to the new location.
_snake_case = os.path.expanduser(
os.getenv("HF_HOME", os.path.join(os.getenv("XDG_CACHE_HOME", "~/.cache"), "huggingface"))
)
_snake_case = os.path.join(hf_cache_home, "diffusers")
def lowerCAmelCase_ ( snake_case_ = None,snake_case_ = None ):
if new_cache_dir is None:
_A : Union[str, Any] = DIFFUSERS_CACHE
if old_cache_dir is None:
_A : str = old_diffusers_cache
_A : Any = Path(snake_case_ ).expanduser()
_A : Tuple = Path(snake_case_ ).expanduser()
for old_blob_path in old_cache_dir.glob("""**/blobs/*""" ):
if old_blob_path.is_file() and not old_blob_path.is_symlink():
_A : Union[str, Any] = new_cache_dir / old_blob_path.relative_to(snake_case_ )
new_blob_path.parent.mkdir(parents=snake_case_,exist_ok=snake_case_ )
os.replace(snake_case_,snake_case_ )
try:
os.symlink(snake_case_,snake_case_ )
except OSError:
logger.warning(
"""Could not create symlink between old cache and new cache. If you use an older version of diffusers again, files will be re-downloaded.""" )
# At this point, old_cache_dir contains symlinks to the new cache (it can still be used).
_snake_case = os.path.join(DIFFUSERS_CACHE, "version_diffusers_cache.txt")
if not os.path.isfile(cache_version_file):
_snake_case = 0
else:
with open(cache_version_file) as f:
try:
_snake_case = int(f.read())
except ValueError:
_snake_case = 0
if cache_version < 1:
_snake_case = os.path.isdir(old_diffusers_cache) and len(os.listdir(old_diffusers_cache)) > 0
if old_cache_is_not_empty:
logger.warning(
"The cache for model files in Diffusers v0.14.0 has moved to a new location. Moving your "
"existing cached models. This is a one-time operation, you can interrupt it or run it "
"later by calling `diffusers.utils.hub_utils.move_cache()`."
)
try:
move_cache()
except Exception as e:
_snake_case = "\n".join(traceback.format_tb(e.__traceback__))
logger.error(
f"""There was a problem when trying to move your cache:\n\n{trace}\n{e.__class__.__name__}: {e}\n\nPlease """
"file an issue at https://github.com/huggingface/diffusers/issues/new/choose, copy paste this whole "
"message and we will do our best to help."
)
if cache_version < 1:
try:
os.makedirs(DIFFUSERS_CACHE, exist_ok=True)
with open(cache_version_file, "w") as f:
f.write("1")
except Exception:
logger.warning(
f"""There was a problem when trying to write in your cache folder ({DIFFUSERS_CACHE}). Please, ensure """
"the directory exists and can be written to."
)
def lowerCAmelCase_ ( snake_case_,snake_case_ = None ):
if variant is not None:
_A : List[str] = weights_name.split(""".""" )
_A : Dict = splits[:-1] + [variant] + splits[-1:]
_A : Any = """.""".join(snake_case_ )
return weights_name
def lowerCAmelCase_ ( snake_case_,*,
snake_case_,snake_case_,snake_case_,snake_case_,snake_case_,snake_case_,snake_case_,snake_case_,snake_case_,snake_case_,snake_case_=None,):
_A : Tuple = str(snake_case_ )
if os.path.isfile(snake_case_ ):
return pretrained_model_name_or_path
elif os.path.isdir(snake_case_ ):
if os.path.isfile(os.path.join(snake_case_,snake_case_ ) ):
# Load from a PyTorch checkpoint
_A : Tuple = os.path.join(snake_case_,snake_case_ )
return model_file
elif subfolder is not None and os.path.isfile(
os.path.join(snake_case_,snake_case_,snake_case_ ) ):
_A : List[Any] = os.path.join(snake_case_,snake_case_,snake_case_ )
return model_file
else:
raise EnvironmentError(
f'''Error no file named {weights_name} found in directory {pretrained_model_name_or_path}.''' )
else:
# 1. First check if deprecated way of loading from branches is used
if (
revision in DEPRECATED_REVISION_ARGS
and (weights_name == WEIGHTS_NAME or weights_name == SAFETENSORS_WEIGHTS_NAME)
and version.parse(version.parse(snake_case_ ).base_version ) >= version.parse("""0.20.0""" )
):
try:
_A : List[Any] = hf_hub_download(
snake_case_,filename=_add_variant(snake_case_,snake_case_ ),cache_dir=snake_case_,force_download=snake_case_,proxies=snake_case_,resume_download=snake_case_,local_files_only=snake_case_,use_auth_token=snake_case_,user_agent=snake_case_,subfolder=snake_case_,revision=revision or commit_hash,)
warnings.warn(
f'''Loading the variant {revision} from {pretrained_model_name_or_path} via `revision=\'{revision}\'` is deprecated. Loading instead from `revision=\'main\'` with `variant={revision}`. Loading model variants via `revision=\'{revision}\'` will be removed in diffusers v1. Please use `variant=\'{revision}\'` instead.''',snake_case_,)
return model_file
except: # noqa: E722
warnings.warn(
f'''You are loading the variant {revision} from {pretrained_model_name_or_path} via `revision=\'{revision}\'`. This behavior is deprecated and will be removed in diffusers v1. One should use `variant=\'{revision}\'` instead. However, it appears that {pretrained_model_name_or_path} currently does not have a {_add_variant(snake_case_,snake_case_ )} file in the \'main\' branch of {pretrained_model_name_or_path}. \n The Diffusers team and community would be very grateful if you could open an issue: https://github.com/huggingface/diffusers/issues/new with the title \'{pretrained_model_name_or_path} is missing {_add_variant(snake_case_,snake_case_ )}\' so that the correct variant file can be added.''',snake_case_,)
try:
# 2. Load model file as usual
_A : Dict = hf_hub_download(
snake_case_,filename=snake_case_,cache_dir=snake_case_,force_download=snake_case_,proxies=snake_case_,resume_download=snake_case_,local_files_only=snake_case_,use_auth_token=snake_case_,user_agent=snake_case_,subfolder=snake_case_,revision=revision or commit_hash,)
return model_file
except RepositoryNotFoundError:
raise EnvironmentError(
f'''{pretrained_model_name_or_path} is not a local folder and is not a valid model identifier '''
"""listed on 'https://huggingface.co/models'\nIf this is a private repository, make sure to pass a """
"""token having permission to this repo with `use_auth_token` or log in with `huggingface-cli """
"""login`.""" )
except RevisionNotFoundError:
raise EnvironmentError(
f'''{revision} is not a valid git identifier (branch name, tag name or commit id) that exists for '''
"""this model name. Check the model page at """
f'''\'https://huggingface.co/{pretrained_model_name_or_path}\' for available revisions.''' )
except EntryNotFoundError:
raise EnvironmentError(
f'''{pretrained_model_name_or_path} does not appear to have a file named {weights_name}.''' )
except HTTPError as err:
raise EnvironmentError(
f'''There was a specific connection error when trying to load {pretrained_model_name_or_path}:\n{err}''' )
except ValueError:
raise EnvironmentError(
f'''We couldn\'t connect to \'{HUGGINGFACE_CO_RESOLVE_ENDPOINT}\' to load this model, couldn\'t find it'''
f''' in the cached files and it looks like {pretrained_model_name_or_path} is not the path to a'''
f''' directory containing a file named {weights_name} or'''
""" \nCheckout your internet connection or see how to run the library in"""
""" offline mode at 'https://huggingface.co/docs/diffusers/installation#offline-mode'.""" )
except EnvironmentError:
raise EnvironmentError(
f'''Can\'t load the model for \'{pretrained_model_name_or_path}\'. If you were trying to load it from '''
"""'https://huggingface.co/models', make sure you don't have a local directory with the same name. """
f'''Otherwise, make sure \'{pretrained_model_name_or_path}\' is the correct path to a directory '''
f'''containing a file named {weights_name}''' )
| 26 |
from collections import OrderedDict
from typing import Mapping
from ...configuration_utils import PretrainedConfig
from ...onnx import OnnxConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"facebook/xmod-base": "https://huggingface.co/facebook/xmod-base/resolve/main/config.json",
"facebook/xmod-large-prenorm": "https://huggingface.co/facebook/xmod-large-prenorm/resolve/main/config.json",
"facebook/xmod-base-13-125k": "https://huggingface.co/facebook/xmod-base-13-125k/resolve/main/config.json",
"facebook/xmod-base-30-125k": "https://huggingface.co/facebook/xmod-base-30-125k/resolve/main/config.json",
"facebook/xmod-base-30-195k": "https://huggingface.co/facebook/xmod-base-30-195k/resolve/main/config.json",
"facebook/xmod-base-60-125k": "https://huggingface.co/facebook/xmod-base-60-125k/resolve/main/config.json",
"facebook/xmod-base-60-265k": "https://huggingface.co/facebook/xmod-base-60-265k/resolve/main/config.json",
"facebook/xmod-base-75-125k": "https://huggingface.co/facebook/xmod-base-75-125k/resolve/main/config.json",
"facebook/xmod-base-75-269k": "https://huggingface.co/facebook/xmod-base-75-269k/resolve/main/config.json",
}
class lowercase ( UpperCamelCase__ ):
_a = "xmod"
def __init__( self , _a=3_0522 , _a=768 , _a=12 , _a=12 , _a=3072 , _a="gelu" , _a=0.1 , _a=0.1 , _a=512 , _a=2 , _a=0.02 , _a=1e-12 , _a=1 , _a=0 , _a=2 , _a="absolute" , _a=True , _a=None , _a=False , _a=2 , _a=False , _a=True , _a=True , _a=("en_XX",) , _a=None , **_a , ) -> str:
super().__init__(pad_token_id=_a , bos_token_id=_a , eos_token_id=_a , **_a )
_A : Tuple = vocab_size
_A : Union[str, Any] = hidden_size
_A : Dict = num_hidden_layers
_A : Dict = num_attention_heads
_A : List[Any] = hidden_act
_A : Optional[Any] = intermediate_size
_A : Any = hidden_dropout_prob
_A : str = attention_probs_dropout_prob
_A : Dict = max_position_embeddings
_A : Any = type_vocab_size
_A : List[Any] = initializer_range
_A : int = layer_norm_eps
_A : int = position_embedding_type
_A : Any = use_cache
_A : int = classifier_dropout
_A : int = pre_norm
_A : Optional[Any] = adapter_reduction_factor
_A : List[Any] = adapter_layer_norm
_A : Optional[int] = adapter_reuse_layer_norm
_A : Any = ln_before_adapter
_A : Union[str, Any] = list(_a )
_A : List[Any] = default_language
class lowercase ( UpperCamelCase__ ):
@property
def a__ ( self ) -> Mapping[str, Mapping[int, str]]:
if self.task == "multiple-choice":
_A : Dict = {0: """batch""", 1: """choice""", 2: """sequence"""}
else:
_A : Dict = {0: """batch""", 1: """sequence"""}
return OrderedDict(
[
("""input_ids""", dynamic_axis),
("""attention_mask""", dynamic_axis),
] )
| 26 | 1 |
from collections import OrderedDict
from typing import Mapping
from packaging import version
from ...configuration_utils import PretrainedConfig
from ...onnx import OnnxConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"facebook/levit-128S": "https://huggingface.co/facebook/levit-128S/resolve/main/config.json",
# See all LeViT models at https://huggingface.co/models?filter=levit
}
class lowercase ( UpperCamelCase__ ):
_a = "levit"
def __init__( self , _a=224 , _a=3 , _a=3 , _a=2 , _a=1 , _a=16 , _a=[128, 256, 384] , _a=[4, 8, 12] , _a=[4, 4, 4] , _a=[16, 16, 16] , _a=0 , _a=[2, 2, 2] , _a=[2, 2, 2] , _a=0.02 , **_a , ) -> Any:
super().__init__(**_a )
_A : List[Any] = image_size
_A : Union[str, Any] = num_channels
_A : Optional[Any] = kernel_size
_A : Optional[int] = stride
_A : int = padding
_A : Optional[int] = hidden_sizes
_A : List[str] = num_attention_heads
_A : Tuple = depths
_A : Any = key_dim
_A : Optional[Any] = drop_path_rate
_A : Tuple = patch_size
_A : Tuple = attention_ratio
_A : int = mlp_ratio
_A : Any = initializer_range
_A : Tuple = [
["""Subsample""", key_dim[0], hidden_sizes[0] // key_dim[0], 4, 2, 2],
["""Subsample""", key_dim[0], hidden_sizes[1] // key_dim[0], 4, 2, 2],
]
class lowercase ( UpperCamelCase__ ):
_a = version.parse("1.11" )
@property
def a__ ( self ) -> Mapping[str, Mapping[int, str]]:
return OrderedDict(
[
("""pixel_values""", {0: """batch""", 1: """num_channels""", 2: """height""", 3: """width"""}),
] )
@property
def a__ ( self ) -> float:
return 1e-4
| 26 |
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
if n == 0:
return 0
_A : Tuple = float("""-inf""" )
for i in range(1,n + 1 ):
_A : str = max(
snake_case_,prices[i - 1] + naive_cut_rod_recursive(n - i,snake_case_ ) )
return max_revue
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
_A : Dict = [float("""-inf""" ) for _ in range(n + 1 )]
return _top_down_cut_rod_recursive(snake_case_,snake_case_,snake_case_ )
def lowerCAmelCase_ ( snake_case_,snake_case_,snake_case_ ):
if max_rev[n] >= 0:
return max_rev[n]
elif n == 0:
return 0
else:
_A : List[str] = float("""-inf""" )
for i in range(1,n + 1 ):
_A : Optional[Any] = max(
snake_case_,prices[i - 1] + _top_down_cut_rod_recursive(n - i,snake_case_,snake_case_ ),)
_A : Tuple = max_revenue
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
_enforce_args(snake_case_,snake_case_ )
# length(max_rev) = n + 1, to accommodate for the revenue obtainable from a rod of
# length 0.
_A : List[Any] = [float("""-inf""" ) for _ in range(n + 1 )]
_A : Any = 0
for i in range(1,n + 1 ):
_A : Optional[Any] = max_rev[i]
for j in range(1,i + 1 ):
_A : int = max(snake_case_,prices[j - 1] + max_rev[i - j] )
_A : int = max_revenue_i
return max_rev[n]
def lowerCAmelCase_ ( snake_case_,snake_case_ ):
if n < 0:
_A : Optional[Any] = f'''n must be greater than or equal to 0. Got n = {n}'''
raise ValueError(snake_case_ )
if n > len(snake_case_ ):
_A : Any = (
"""Each integral piece of rod must have a corresponding price. """
f'''Got n = {n} but length of prices = {len(snake_case_ )}'''
)
raise ValueError(snake_case_ )
def lowerCAmelCase_ ( ):
_A : Tuple = [6, 10, 12, 15, 20, 23]
_A : List[Any] = len(snake_case_ )
# the best revenue comes from cutting the rod into 6 pieces, each
# of length 1 resulting in a revenue of 6 * 6 = 36.
_A : Any = 36
_A : List[Any] = top_down_cut_rod(snake_case_,snake_case_ )
_A : List[Any] = bottom_up_cut_rod(snake_case_,snake_case_ )
_A : Dict = naive_cut_rod_recursive(snake_case_,snake_case_ )
assert expected_max_revenue == max_rev_top_down
assert max_rev_top_down == max_rev_bottom_up
assert max_rev_bottom_up == max_rev_naive
if __name__ == "__main__":
main()
| 26 | 1 |
from collections import OrderedDict
from typing import Mapping
from ...configuration_utils import PretrainedConfig
from ...onnx import OnnxConfig
from ...utils import logging
_snake_case = logging.get_logger(__name__)
_snake_case = {
"xlm-mlm-en-2048": "https://huggingface.co/xlm-mlm-en-2048/resolve/main/config.json",
"xlm-mlm-ende-1024": "https://huggingface.co/xlm-mlm-ende-1024/resolve/main/config.json",
"xlm-mlm-enfr-1024": "https://huggingface.co/xlm-mlm-enfr-1024/resolve/main/config.json",
"xlm-mlm-enro-1024": "https://huggingface.co/xlm-mlm-enro-1024/resolve/main/config.json",
"xlm-mlm-tlm-xnli15-1024": "https://huggingface.co/xlm-mlm-tlm-xnli15-1024/resolve/main/config.json",
"xlm-mlm-xnli15-1024": "https://huggingface.co/xlm-mlm-xnli15-1024/resolve/main/config.json",
"xlm-clm-enfr-1024": "https://huggingface.co/xlm-clm-enfr-1024/resolve/main/config.json",
"xlm-clm-ende-1024": "https://huggingface.co/xlm-clm-ende-1024/resolve/main/config.json",
"xlm-mlm-17-1280": "https://huggingface.co/xlm-mlm-17-1280/resolve/main/config.json",
"xlm-mlm-100-1280": "https://huggingface.co/xlm-mlm-100-1280/resolve/main/config.json",
}
class lowercase ( UpperCamelCase__ ):
_a = "xlm"
_a = {
"hidden_size": "emb_dim",
"num_attention_heads": "n_heads",
"num_hidden_layers": "n_layers",
"n_words": "vocab_size", # For backward compatibility
}
def __init__( self , _a=3_0145 , _a=2048 , _a=12 , _a=16 , _a=0.1 , _a=0.1 , _a=True , _a=False , _a=False , _a=False , _a=1 , _a=True , _a=512 , _a=2048**-0.5 , _a=1e-12 , _a=0.02 , _a=0 , _a=1 , _a=2 , _a=3 , _a=5 , _a=True , _a="first" , _a=True , _a=None , _a=True , _a=0.1 , _a=5 , _a=5 , _a=0 , _a=0 , _a=2 , _a=0 , **_a , ) -> Optional[int]:
_A : Optional[int] = vocab_size
_A : Optional[Any] = emb_dim
_A : Optional[int] = n_layers
_A : Optional[int] = n_heads
_A : List[str] = dropout
_A : Optional[int] = attention_dropout
_A : Optional[Any] = gelu_activation
_A : Union[str, Any] = sinusoidal_embeddings
_A : Union[str, Any] = causal
_A : List[str] = asm
_A : int = n_langs
_A : List[Any] = use_lang_emb
_A : Any = layer_norm_eps
_A : str = bos_index
_A : Union[str, Any] = eos_index
_A : Optional[Any] = pad_index
_A : Optional[int] = unk_index
_A : str = mask_index
_A : Tuple = is_encoder
_A : Dict = max_position_embeddings
_A : Tuple = embed_init_std
_A : Optional[Any] = init_std
_A : Tuple = summary_type
_A : Optional[int] = summary_use_proj
_A : Optional[Any] = summary_activation
_A : Dict = summary_proj_to_labels
_A : Union[str, Any] = summary_first_dropout
_A : Tuple = start_n_top
_A : int = end_n_top
_A : Optional[Any] = mask_token_id
_A : Union[str, Any] = lang_id
if "n_words" in kwargs:
_A : List[str] = kwargs["""n_words"""]
super().__init__(pad_token_id=_a , bos_token_id=_a , **_a )
class lowercase ( UpperCamelCase__ ):
@property
def a__ ( self ) -> Mapping[str, Mapping[int, str]]:
if self.task == "multiple-choice":
_A : Any = {0: """batch""", 1: """choice""", 2: """sequence"""}
else:
_A : str = {0: """batch""", 1: """sequence"""}
return OrderedDict(
[
("""input_ids""", dynamic_axis),
("""attention_mask""", dynamic_axis),
("""token_type_ids""", dynamic_axis),
] )
| 26 |
import requests
from bsa import BeautifulSoup
def lowerCAmelCase_ ( snake_case_ = "AAPL" ):
_A : str = f'''https://in.finance.yahoo.com/quote/{symbol}?s={symbol}'''
_A : List[Any] = BeautifulSoup(requests.get(snake_case_ ).text,"""html.parser""" )
_A : Union[str, Any] = """My(6px) Pos(r) smartphone_Mt(6px)"""
return soup.find("""div""",class_=class_ ).find("""span""" ).text
if __name__ == "__main__":
for symbol in "AAPL AMZN IBM GOOG MSFT ORCL".split():
print(f"""Current {symbol:<4} stock price is {stock_price(symbol):>8}""")
| 26 | 1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.