import streamlit as st import os #from transformers import BertModel, BertTokenizer from transformers import HfAgent, load_tool import torch from transformers import AutoModelForCausalLM, AutoTokenizer, LocalAgent #checkpoint = "THUDM/agentlm-7b" #model = AutoModelForCausalLM.from_pretrained(checkpoint, device_map="auto", torch_dtype=torch.bfloat16) #tokenizer = AutoTokenizer.from_pretrained(checkpoint) #agent = LocalAgent(model, tokenizer) #agent.run("Draw me a picture of rivers and lakes.") #print(agent.run("Is the following `text` (in Spanish) positive or negative?", text="¡Este es un API muy agradable!")) # Load tools controlnet_transformer = load_tool("huggingface-tools/text-to-image") upscaler = load_tool("diffusers/latent-upscaler-tool") tools = [controlnet_transformer, upscaler ] ############ HfAgent from huggingface_hub import login #Do this before HfAgent() and it should work #from huggingface_hub import login # load tools from transformers.tools import HfAgent from transformers.tools import Agent #import textract #from utils import logging import time from huggingface_hub import HfFolder, hf_hub_download, list_spaces class CustomHfAgent(Agent): """ Agent that uses an inference endpoint to generate code. Args: url_endpoint (`str`): The name of the url endpoint to use. token (`str`, *optional*): The token to use as HTTP bearer authorization for remote files. If unset, will use the token generated when running `huggingface-cli login` (stored in `~/.huggingface`). chat_prompt_template (`str`, *optional*): Pass along your own prompt if you want to override the default template for the `chat` method. Can be the actual prompt template or a repo ID (on the Hugging Face Hub). The prompt should be in a file named `chat_prompt_template.txt` in this repo in this case. run_prompt_template (`str`, *optional*): Pass along your own prompt if you want to override the default template for the `run` method. Can be the actual prompt template or a repo ID (on the Hugging Face Hub). The prompt should be in a file named `run_prompt_template.txt` in this repo in this case. additional_tools ([`Tool`], list of tools or dictionary with tool values, *optional*): Any additional tools to include on top of the default ones. If you pass along a tool with the same name as one of the default tools, that default tool will be overridden. Example: ```py from transformers import HfAgent agent = HfAgent("https://api-inference.huggingface.co/models/bigcode/starcoder") agent.run("Is the following `text` (in Spanish) positive or negative?", text="¡Este es un API muy agradable!") ``` """ def __init__( self, url_endpoint, token=os.environ['HF_token'], chat_prompt_template=None, run_prompt_template=None, additional_tools=None ): # super()._init_(self, url_endpoint, token=None, chat_prompt_template=None, run_prompt_template=None, additional_tools=None) self.url_endpoint = url_endpoint if token is None: self.token = f"Bearer {HfFolder().get_token()}" elif token.startswith("Bearer") or token.startswith("Basic"): self.token = token else: self.token = f"Bearer {token}" super().__init__( chat_prompt_template=chat_prompt_template, run_prompt_template=run_prompt_template, additional_tools=additional_tools, ) def generate_one(self, prompt, stop): headers = {"Authorization": self.token} inputs = { "inputs": prompt, "parameters": {"max_new_tokens": 192, "return_full_text": False, "stop": stop}, } print(inputs) response = requests.post(self.url_endpoint, json=inputs, headers=headers) if response.status_code == 429: print("Getting rate-limited, waiting a tiny bit before trying again.") time.sleep(1) return self._generate_one(prompt) elif response.status_code != 200: raise ValueError(f"Errors {inputs} {response.status_code}: {response.json()}") result = response.json()[0]["generated_text"] # Inference API returns the stop sequence for stop_seq in stop: if result.endswith(stop_seq): return result[: -len(stop_seq)] return result # create agent #agent = HfAgent(API_URL) #print(agent) # instruct agent # Use CustomHfAgent in your code agent = CustomHfAgent("https://api-inference.huggingface.co/models/bigcode/starcoder") print("-----") print(agent.token) print("-----") agent.token = os.environ['HF_token'] print("-----") print(agent.token) print("-----") #agent.token = "Bearer xxx" #print(agent.token) #agent.run("Answer the following question", question ="what is the capitol of the usa?", context="The capitol of the usa is London") agent.chat("Draw me a picture of rivers and lakes") #agent.chat("Transform the picture so that there is a rock in there") #result = agent.generate_one("What is the capitol of the usa.", stop=["your_stop_sequence"]) #print(result) #agent.run("Show me an image of a horse") ##### # Define the model and tokenizer #model = BertModel.from_pretrained('bert-base-uncased') #tokenizer = BertTokenizer.from_pretrained('bert-base-uncased') # Create the Streamlit app st.title("Hugging Face Agent") # Input field for the user's message message_input = st.text_input("Enter your message:", "") # Checkboxes for the tools to be used by the agent tool_checkboxes = [st.checkbox(f"Use {tool}") for tool in tools] # Submit button #submit_button = st.button("Submit") # Define the callback function to handle the form submission def handle_submission(): # Get the user's message and the selected tools message = message_input selected_tools = [tool for tool, checkbox in zip(tools, tool_checkboxes) if checkbox] # Initialize the agent with the selected tools #agent = HfAgent("https://api-inference.huggingface.co/models/bigcode/starcoder", additional_tools=tools) #agent = HfAgent("https://api-inference.huggingface.co/models/OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5", additional_tools=tools) agent = HfAgent("https://api-inference.huggingface.co/models/THUDM/agentlm-7b", additional_tools=tools) # agent.config.tokenizer = tokenizer # agent.config.tools = selected_tools # Process the user's message # inputs = tokenizer.encode_plus(message, add_special_tokens=True, return_tensors="pt") # outputs = agent(inputs['input_ids'], attention_mask=inputs['attention_mask']) # Display the agent's response response = agent.run(message) st.text(f"{response:.4f}") return "done" # Add the callback function to the Streamlit app submit_button = st.button("Submit", on_click=handle_submission)