zac's picture
Update app.py
117600f
raw
history blame
930 Bytes
import gradio as gr
from huggingface_hub import hf_hub_download
from llama_cpp import Llama
llm = Llama(model_path=hf_hub_download(repo_id="TheBloke/airoboros-l2-13b-gpt4-m2.0-GGML", filename="airoboros-l2-13b-gpt4-m2.0.ggmlv3.q6_K.bin"), n_ctx=2048)
history = []
history.append(["Hi there!", "Hello, how can I help you?"])
def generate_text(input_text, history):
conversation_context = " ".join([f"{pair[0]} {pair[1]}" for pair in history])
full_conversation = f"{conversation_context} Q: {input_text} \n A:"
output = llm(full_conversation, max_tokens=1024, stop=["Q:", "\n"], echo=True)
response = output['choices'][0]['text']
history.append([input_text, response])
return "", history
with gr.Blocks() as demo:
chatbot = gr.Chatbot()
msg = gr.Textbox()
clear = gr.ClearButton([msg, chatbot])
msg.submit(generate_text, [msg, chatbot], [msg, chatbot])
demo.launch()