app.py

import gradio as gr
import requests

def respond(message, history, system_message, max_tokens, temperature, top_p):
    messages = [{"role": "system", "content": system_message}]

    for val in history:
        if val[0]:
            messages.append({"role": "user", "content": val[0]})
        if val[1]:
            messages.append({"role": "assistant", "content": val[1]})

    messages.append({"role": "user", "content": message})

    response = ""

    data = {
        "model": "hf.co/ibrahimBlyc/LA_Llama:latest",
        "prompt": "", # On construit le prompt complet ci-dessous
        "max_tokens": max_tokens,
        "temperature": temperature,
        "top_p": top_p
        # Autres paramètres Ollama si nécessaire
    }

    # Construire le prompt complet à partir des messages
    prompt = ""
    for msg in messages:
        prompt += f"{msg['role']}: {msg['content']}\n"
    data["prompt"] = prompt

    url = "http://localhost:11434/api/generate"
    
    try:
        stream_response = requests.post(url, json=data, stream=True)
        stream_response.raise_for_status() # Lève une exception si le code de status n'est pas 2xx

        for chunk in stream_response.iter_lines():
            if chunk:
                decoded_chunk = chunk.decode()
                try:
                    response_json = eval(decoded_chunk) # Évalue la réponse JSON.  Attention à la sécurité !
                    token = response_json.get("response", "")

                    if token:
                        response += token
                        yield response
                except (SyntaxError, NameError, json.JSONDecodeError) as e:
                    print(f"Erreur lors du décodage du chunk : {e}. Chunk : {decoded_chunk}")
                    yield f"Erreur: Impossible de décoder la réponse du serveur."
                    return

    except requests.exceptions.RequestException as e:
        print(f"Erreur de requête : {e}")
        yield f"Erreur: Impossible de communiquer avec le serveur Ollama."


demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (nucleus sampling)",
        ),
    ],
)


if __name__ == "__main__":
    demo.launch(share=True)