Spaces:
Sleeping
Sleeping
| from openai import OpenAI | |
| import gradio as gr | |
| import time | |
| anyscale_base_url = "https://api.endpoints.anyscale.com/v1" | |
| def predict(api_key, user_input): | |
| prompt = f"""[INST] {user_input} [/INST]""" | |
| client = OpenAI(base_url=anyscale_base_url, api_key=api_key) | |
| completion = client.completions.create( | |
| model="mistralai/Mixtral-8x7B-Instruct-v0.1", | |
| prompt=prompt, | |
| temperature=0.7, | |
| max_tokens=100) | |
| response = completion.choices[0].text | |
| # Simulate streaming by splitting the response and yielding parts with delays | |
| words = response.split() | |
| for i in range(0, len(words), 5): # Adjust the step to control chunk sizes | |
| yield ' '.join(words[i:i+5]) | |
| time.sleep(1) # Adjust the sleep time to control the streaming speeded | |
| #return response | |
| def main(): | |
| description = "This is a simple interface to interact with OpenAI’s Chat Completion API. Please enter your API key and your message." | |
| with gr.Blocks() as demo: | |
| with gr.Row(): | |
| api_key_input = gr.Textbox(label="API Key", placeholder="Enter your API key here", show_label=True, type="password") | |
| user_input = gr.Textbox(label="Your Message", placeholder="Enter your message here") | |
| submit_btn = gr.Button("Submit") | |
| output = gr.Textbox(label="LLM Response", live=True) | |
| #output = gr.Textbox(label="LLM Response") | |
| submit_btn.click(fn=predict, inputs=[api_key_input, user_input], outputs=output) | |
| demo.launch() | |
| if __name__ == "__main__": | |
| main() | |