app.py

import os
import gradio as gr
from huggingface_hub import InferenceClient
from datasets import load_dataset
import time
import faiss
import numpy as np

# ✅ Install FAISS if missing
os.system("pip install faiss-cpu")

def log(message):
    print(f"✅ {message}")


# ✅ Load the datasets
datasets = {
    "sales": load_dataset("goendalf666/sales-conversations", trust_remote_code=True),
    "blended": load_dataset("blended_skill_talk", trust_remote_code=True),
    "dialog": load_dataset("daily_dialog", trust_remote_code=True),
    "multiwoz": load_dataset("multi_woz_v22", trust_remote_code=True),
}

# Optional: Print dataset names and sizes
for name, dataset in datasets.items():
    print(f"{name}: {len(dataset['train'])} examples")

# Initialize the model client (use correct model for chatbot)
client = InferenceClient("mistralai/Mistral-7B-Instruct-v0.3")

# Chatbot response function
def respond(
    message,
    history: list[tuple[str, str]],
    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 = ""

    for message in client.chat_completions(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        token = message["choices"][0]["delta"]["content"]
        response += token
        yield response


# Gradio interface for chatbot
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)",
        ),
    ],
)

def start_embedding():
    # Include your embedding logic here (from embeddings.py)
    log("Embedding started...")
    time.sleep(2)  # Simulating embedding process
    log("Embedding process finished.")

# Create Gradio interface with a button to start the embedding
demo = gr.Interface(
    fn=start_embedding,
    inputs=None,
    outputs="text",
    live=True,
    title="Embedding Trigger"
)


# ✅ Function to check FAISS index
def check_faiss():
    index_path = "my_embeddings"  # Adjust if needed

    try:
        index = faiss.read_index(index_path)
        num_vectors = index.ntotal
        dim = index.d

        if num_vectors > 0:
            sample_vectors = index.reconstruct_n(0, min(5, num_vectors))  # Get first 5 embeddings
            return f"📊 FAISS index contains {num_vectors} vectors.\n✅ Embedding dimension: {dim}\n🧐 Sample: {sample_vectors[:2]} ..."
        else:
            return "⚠️ No embeddings found in FAISS index!"

    except Exception as e:
        return f"❌ ERROR: Failed to load FAISS index - {e}"

# ✅ Add a Gradio button to trigger FAISS check
with gr.Blocks() as demo:
    gr.Markdown("### 🔍 FAISS Embedding Check")
    
    check_button = gr.Button("🔎 Check FAISS Embeddings")
    output_text = gr.Textbox(label="FAISS Status", interactive=False)

    check_button.click(fn=check_faiss, outputs=output_text)

# Launch Gradio app
if __name__ == "__main__":
    demo.launch()