Taf2023 commited on
Commit
79a8fa3
·
verified ·
1 Parent(s): 573f399

Upload app (4).py

Browse files
Files changed (1) hide show
  1. app (4).py +1 -0
app (4).py ADDED
@@ -0,0 +1 @@
 
 
1
+ from huggingface_hub import InferenceClient import gradio as gr import random import prompts client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1") def format_prompt(message, history): prompt = "<s>" for user_prompt, bot_response in history: prompt += f"[INST] {user_prompt} [/INST]" prompt += f" {bot_response}</s> " prompt += f"[INST] {message} [/INST]" return prompt def generate(prompt, history, max_new_tokens, seed): print(seed) system_prompt = prompts.ARTICLE_WRITER.format(history=history) generate_kwargs = dict( temperature=0.9, max_new_tokens=max_new_tokens, top_p=0.95, repetition_penalty=1.0, do_sample=True, seed=seed, ) formatted_prompt = format_prompt(f"{system_prompt}, {prompt}", history) stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False) output = "" for response in stream: output += response.token.text yield output additional_inputs = [ gr.Slider( label="Max new tokens", value=4096, minimum=10, maximum=1048*10, step=64, interactive=True, info="The maximum numbers of words the chatbot will return", ), gr.Slider( label="Seed", value=random.randint(1,9999999999999999), minimum=0, maximum=9999999999999999, step=64, interactive=True, info="Each seed produces a different output to a single prompt", ), ] examples = [ ["Write a 1000-word article about the benefits of a healthy lifestyle."], ["Compose a 2000-word essay on the impact of technology on modern society."], ["Generate a 1500-word article discussing the importance of environmental conservation."], ["Create a 3000-word piece exploring the role of art in shaping cultural identity."], ] gr.ChatInterface( fn=generate, chatbot=gr.Chatbot(show_label=False, show_share_button=False, show_copy_button=True, likeable=True, layout="panel"), title="Article Writer", additional_inputs=additional_inputs, examples=examples, concurrency_limit=20, ).launch(show_api=False)