Spaces:
Running
Running
| import os | |
| import gradio as gr | |
| title = "Have Fun With ChubbyBot" | |
| description = """ | |
| <p> | |
| <center> | |
| The bot is trained on blended_skill_talk dataset using facebook/blenderbot-400M-distill. | |
| <img src="https://huggingface.co/spaces/EXFINITE/BlenderBot-UI/resolve/main/img/cover.png" alt="rick" width="250"/> | |
| </center> | |
| </p> | |
| """ | |
| article = "<p style='text-align: center'><a href='https://arxiv.org/abs/1907.06616' target='_blank'>Recipes for building an open-domain chatbot</a></p><p style='text-align: center'><a href='https://parl.ai/projects/recipes/' target='_blank'>Original PARLAI Code</a></p></center></p>" | |
| import torch | |
| from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, BlenderbotForConditionalGeneration, BlenderbotForCausalLM, BlenderbotTokenizer | |
| tokenizer = BlenderbotTokenizer.from_pretrained("facebook/blenderbot-400M-distill") | |
| model = BlenderbotForConditionalGeneration.from_pretrained("facebook/blenderbot-400M-distill",add_cross_attention=False) | |
| def predict(input, history=[]): | |
| # tokenize the new input sentence | |
| new_user_input_ids = tokenizer.encode(input + tokenizer.eos_token, return_tensors='pt') | |
| # append the new user input tokens to the chat history | |
| bot_input_ids = torch.cat([torch.LongTensor(history), new_user_input_ids], dim=-1) | |
| # generate a response | |
| history = model.generate(bot_input_ids, max_length=1000, pad_token_id=tokenizer.eos_token_id).tolist() | |
| # convert the tokens to text, and then split the responses into the right format | |
| response = tokenizer.decode(history[0]).replace("<s>","").split("</s>") | |
| response = [(response[i], response[i+1]) for i in range(0, len(response), 2)] # convert to tuples of list | |
| return response, history | |
| gr.Interface( | |
| fn = predict, | |
| inputs = ["textbox","state"], | |
| outputs = ["chatbot","state"], | |
| theme ="seafoam", | |
| title = title, | |
| description = description, | |
| article = article | |
| ).launch(enable_queue=True) | |