Elisa / app.py
faelfernandes's picture
Update app.py
2994fcb verified
from huggingface_hub import InferenceClient
import gradio as gr
client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")
def format_prompt(message, history, system_prompt=None):
prompt = "<s>"
for user_prompt, bot_response in history:
prompt += f"[INST] {user_prompt} [/INST]"
prompt += f" {bot_response}</s> "
if system_prompt:
prompt += f"[SYS] {system_prompt} [/SYS]"
prompt += f"[INST] {message} [/INST]"
return prompt
def generate(
prompt, history, system_prompt="Sou o assistente TSM, pronto para auxiliar em qualquer tarefa, meu idioma principal é o português brasileiro", temperature=0.2, max_new_tokens=8192, top_p=0.95, repetition_penalty=1.0,
):
temperature = float(temperature)
if temperature < 1e-2:
temperature = 1e-2
top_p = float(top_p)
generate_kwargs = dict(
temperature=temperature,
max_new_tokens=max_new_tokens,
top_p=top_p,
repetition_penalty=repetition_penalty,
do_sample=True,
seed=42,
)
formatted_prompt = format_prompt(prompt, history, system_prompt)
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
return output
mychatbot = gr.Chatbot(
avatar_images=["./user.png", "./botm.png"], bubble_full_width=False, show_label=False, show_copy_button=True, likeable=True,)
demo = gr.ChatInterface(
fn=generate,
chatbot=mychatbot,
title="Olá, eu sou o TSM 3.5 em que posso ajudar?",
css="body { background-color: inherit; overflow-x:hidden;}"
":root {--color-accent: transparent !important; --color-accent-soft:transparent !important; --code-background-fill:black !important; --body-text-color:white !important;}"
"#component-2 {background:#ffffff1a; display:contents;}"
"div#component-0 { height: auto !important;}"
".gradio-container.gradio-container-4-8-0.svelte-1kyws56.app {background: #000000;}gradio-app {background: #333 !important;}"
"gradio-app {background: #000000 !important; background-attachment: fixed !important; background-position: top;}"
".panel.svelte-vt1mxs {background: transparent; padding:0;}"
".block.svelte-90oupt { background: transparent; border-color: transparent;}"
".bot.svelte-12dsd9j.svelte-12dsd9j.svelte-12dsd9j { background: #ffffff1a; border-color: transparent; color: white;}"
".user.svelte-12dsd9j.svelte-12dsd9j.svelte-12dsd9j { background: #ffffff1a; border-color: transparent; color: white; padding: 10px 18px;}"
"div.svelte-iyf88w{ background: #cc0000; border-color: trasparent; border-radius: 25px;}"
"textarea.scroll-hide.svelte-1f354aw {background: #555 !important;}"
".primary.svelte-cmf5ev { background: transparent; color: white;}"
".primary.svelte-cmf5ev:hover { background: transparent; color: white;}"
"div#component-9 { max-width: fit-content; margin-left: auto; margin-right: auto;}"
"button#component-8, button#component-10, button#component-11, button#component-12 { flex: none; background: #ffffff1a; border: none; color: white; margin-right: auto; margin-left: auto; border-radius: 9px; min-width: fit-content;}"
".share-button.svelte-12dsd9j { display: none;}"
"footer.svelte-mpyp5e { display: none !important;}"
".message-buttons-bubble.svelte-12dsd9j.svelte-12dsd9j.svelte-12dsd9j { border-color: #FFFFFF; background: #FFFFFF;}"
".bubble-wrap.svelte-12dsd9j.svelte-12dsd9j.svelte-12dsd9j {padding: 0;}"
".prose h1 { color: white !important; font-size: 16px !important; font-weight: normal !important; background: #ffffff1a; padding: 20px; border-radius: 20px; width: 90%; margin-left: auto !important; margin-right: auto !important;}"
".toast-wrap.svelte-pu0yf1 { display:none !important;}"
".scroll-hide { scrollbar-width: auto !important;}"
".main svelte-1kyws56 {max-width: 800px; align-self: center;}"
"div#component-4 {max-width: 650px; margin-left: auto; margin-right: auto;}"
"body::-webkit-scrollbar { display: none;}"
)
demo.queue().launch(show_api=False)