google-gemma / app.py
Omnibus's picture
Create app.py
04c25c5 verified
raw
history blame
1.57 kB
import gradio as gr
from huggingface_hub import InferenceClient
client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")
def chat_inf(system_prompt,prompt,history):
if not history:
history = []
hist_len=0
if history:
hist_len=len(history)
print(hist_len)
seed = random.randint(1,1111111111111111)
generate_kwargs = dict(
temperature=0.9,
max_new_tokens=10480,
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 [(prompt,output)]
history.append((prompt,output))
yield history
with gr.Blocks() as app:
with gr.Group():
chat_b = gr.Chatbot()
with gr.Row():
with gr.Column(scale=3):
inp = gr.Textbox(label="Prompt")
sys_inp = gr.Textbox(label="System Prompt (optional)")
btn = gr.Button("Chat")
with gr.Column(scale=1):
with gr.Group():
stop_btn=gr.Button("Stop")
clear_btn=gr.Button("Clear")
chatblock=gr.Dropdown(label="Chatblocks",choices=[c for c in range(1,40)],multiselect=True)
btn.click(chat_inf,[sys_inp,inp,chat_b],chat_b)
app.launch()