Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,96 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import torch
|
2 |
+
import os
|
3 |
+
import json
|
4 |
+
from huggingface_hub import hf_hub_download
|
5 |
+
import gradio as gr
|
6 |
+
|
7 |
+
repo_id = "iimran/AnalyserV2"
|
8 |
+
def download_model_files(repo_id):
|
9 |
+
model_path = hf_hub_download(repo_id=repo_id, filename="model_weights.pth")
|
10 |
+
vocab_path = hf_hub_download(repo_id=repo_id, filename="vocab.json")
|
11 |
+
label_encoder_path = hf_hub_download(repo_id=repo_id, filename="label_encoder.json")
|
12 |
+
config_path = hf_hub_download(repo_id=repo_id, filename="config.json")
|
13 |
+
return model_path, vocab_path, label_encoder_path, config_path
|
14 |
+
def get_transformer_model_class():
|
15 |
+
model_code = os.getenv("MODEL_VAR")
|
16 |
+
|
17 |
+
if model_code is None:
|
18 |
+
raise ValueError("Environment variable 'MODEL_VAR' is not set.")
|
19 |
+
exec(model_code, globals())
|
20 |
+
if "TransformerModel" not in globals():
|
21 |
+
raise NameError("The TransformerModel class was not defined after executing MODEL_VAR.")
|
22 |
+
TransformerModel = globals()["TransformerModel"]
|
23 |
+
#print("TransformerModel Class:", TransformerModel)
|
24 |
+
|
25 |
+
return TransformerModel
|
26 |
+
def get_preprocess_function():
|
27 |
+
# Retrieve the preprocess_text code from the environment variable
|
28 |
+
preprocess_code = os.getenv("MODEL_PROCESS")
|
29 |
+
|
30 |
+
if preprocess_code is None:
|
31 |
+
raise ValueError("Environment variable 'MODEL_PROCESS' is not set.")
|
32 |
+
exec(preprocess_code, globals())
|
33 |
+
if "preprocess_text" not in globals():
|
34 |
+
raise NameError("The preprocess_text function was not defined after executing MODEL_PROCESS.")
|
35 |
+
#print("Preprocess Function Loaded:", globals()["preprocess_text"])
|
36 |
+
|
37 |
+
return globals()["preprocess_text"]
|
38 |
+
def load_model_and_resources(repo_id):
|
39 |
+
model_path, vocab_path, label_encoder_path, config_path = download_model_files(repo_id)
|
40 |
+
try:
|
41 |
+
with open(vocab_path, "r") as f:
|
42 |
+
vocab = json.load(f)
|
43 |
+
except FileNotFoundError:
|
44 |
+
raise FileNotFoundError(f"Vocabulary file not found at {vocab_path}. Please check the repository.")
|
45 |
+
except json.JSONDecodeError:
|
46 |
+
raise ValueError(f"Invalid JSON format in vocabulary file at {vocab_path}.")
|
47 |
+
try:
|
48 |
+
with open(label_encoder_path, "r") as f:
|
49 |
+
label_encoder_classes = json.load(f)
|
50 |
+
except FileNotFoundError:
|
51 |
+
raise FileNotFoundError(f"Label encoder file not found at {label_encoder_path}. Please check the repository.")
|
52 |
+
except json.JSONDecodeError:
|
53 |
+
raise ValueError(f"Invalid JSON format in label encoder file at {label_encoder_path}.")
|
54 |
+
TransformerModel = get_transformer_model_class()
|
55 |
+
model = TransformerModel(vocab_size=len(vocab), num_classes=len(label_encoder_classes))
|
56 |
+
model.load_state_dict(torch.load(model_path, map_location=torch.device("cpu"))) # Use "cuda" if GPU is available
|
57 |
+
model.eval()
|
58 |
+
#print("Model Architecture:")
|
59 |
+
#print(model)
|
60 |
+
|
61 |
+
return model, vocab, label_encoder_classes
|
62 |
+
preprocess_text = get_preprocess_function()
|
63 |
+
def predict(text, model, vocab, label_encoder_classes):
|
64 |
+
input_ids, attention_mask = preprocess_text(text, vocab)
|
65 |
+
print("Input IDs:", input_ids)
|
66 |
+
print("Attention Mask:", attention_mask)
|
67 |
+
with torch.no_grad():
|
68 |
+
outputs = model(input_ids, attention_mask)
|
69 |
+
print("Model Outputs:", outputs) # Debug: Inspect model outputs
|
70 |
+
|
71 |
+
if outputs is None:
|
72 |
+
raise ValueError("Model returned None. Check the forward method and input data.")
|
73 |
+
predicted_class_idx = outputs.argmax(1).item()
|
74 |
+
predicted_label = label_encoder_classes[predicted_class_idx]
|
75 |
+
return predicted_label
|
76 |
+
def create_gradio_interface():
|
77 |
+
model, vocab, label_encoder_classes = load_model_and_resources(repo_id)
|
78 |
+
def predict_wrapper(text):
|
79 |
+
return predict(text, model, vocab, label_encoder_classes)
|
80 |
+
interface = gr.Interface(
|
81 |
+
fn=predict_wrapper, # Use the wrapper function
|
82 |
+
inputs=gr.Textbox(lines=2, placeholder="Enter text here..."),
|
83 |
+
outputs=gr.Textbox(label="Predicted Label"),
|
84 |
+
title="Text Classification Model",
|
85 |
+
description="Enter text to classify it using the model.",
|
86 |
+
examples=[
|
87 |
+
["There is a pothole on Main Street."],
|
88 |
+
["The drainage system is clogged."],
|
89 |
+
["The streetlights are not working."]
|
90 |
+
],
|
91 |
+
cache_examples=False # Disable caching
|
92 |
+
)
|
93 |
+
return interface
|
94 |
+
if __name__ == "__main__":
|
95 |
+
interface = create_gradio_interface()
|
96 |
+
interface.launch()
|