import torch import gradio as gr from transformers import pipeline, AutoModelForSeq2SeqLM, AutoTokenizer, AutoConfig # List of summarization models model_names = [ "google/bigbird-pegasus-large-arxiv", "facebook/bart-large-cnn", "google/t5-v1_1-large", "sshleifer/distilbart-cnn-12-6", "allenai/led-base-16384", "google/pegasus-xsum", "togethercomputer/LLaMA-2-7B-32K" ] # Placeholder for the summarizer pipeline, tokenizer, and maximum tokens summarizer = None tokenizer = None max_tokens = None # Function to load the selected model def load_model(model_name): global summarizer, tokenizer, max_tokens try: # Load the summarization pipeline with the selected model summarizer = pipeline("summarization", model=model_name, torch_dtype=torch.bfloat16) # Load the tokenizer for the selected model tokenizer = AutoTokenizer.from_pretrained(model_name) # Load the configuration for the selected model config = AutoConfig.from_pretrained(model_name) # Determine the maximum tokens based on available configuration attributes if hasattr(config, 'max_position_embeddings'): max_tokens = config.max_position_embeddings elif hasattr(config, 'n_positions'): max_tokens = config.n_positions elif hasattr(config, 'd_model'): max_tokens = config.d_model # for T5 models, d_model is a rough proxy else: max_tokens = "Unknown" return f"Model {model_name} loaded successfully! Max tokens: {max_tokens}" except Exception as e: return f"Failed to load model {model_name}. Error: {str(e)}" # Function to summarize the input text def summarize_text(input, min_length, max_length): if summarizer is None: return "No model loaded!" # Tokenize the input text and check the number of tokens input_tokens = tokenizer.encode(input, return_tensors="pt") num_tokens = input_tokens.shape[1] if num_tokens > max_tokens: # Return an error message if the input text exceeds the maximum token limit return f"Error: The input text has {num_tokens} tokens, which exceeds the maximum allowed {max_tokens} tokens. Please enter shorter text." # Calculate minimum and maximum summary length based on the percentages min_summary_length = int(num_tokens * (min_length / 100)) max_summary_length = int(num_tokens * (max_length / 100)) # Summarize the input text using the loaded model with specified lengths output = summarizer(input, min_length=min_summary_length, max_length=max_summary_length) return output[0]['summary_text'] # Gradio Interface with gr.Blocks() as demo: with gr.Row(): # Dropdown menu for selecting the model model_dropdown = gr.Dropdown(choices=model_names, label="Choose a model", value="sshleifer/distilbart-cnn-12-6") # Button to load the selected model load_button = gr.Button("Load Model") # Textbox to display the load status load_message = gr.Textbox(label="Load Status", interactive=False) # Slider for minimum summary length min_length_slider = gr.Slider(minimum=0, maximum=100, step=1, label="Minimum Summary Length (%)", value=10) # Slider for maximum summary length max_length_slider = gr.Slider(minimum=0, maximum=100, step=1, label="Maximum Summary Length (%)", value=20) # Textbox for inputting the text to be summarized input_text = gr.Textbox(label="Input text to summarize", lines=6) # Button to trigger the summarization summarize_button = gr.Button("Summarize Text") # Textbox to display the summarized text output_text = gr.Textbox(label="Summarized text", lines=4) # Define the actions for the load button and summarize button load_button.click(fn=load_model, inputs=model_dropdown, outputs=load_message) summarize_button.click(fn=summarize_text, inputs=[input_text, min_length_slider, max_length_slider], outputs=output_text) # Launch the Gradio interface demo.launch()