app.py

from flask import Flask, request, jsonify
from flask_cors import CORS
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

app = Flask(__name__)
CORS(app)

# Model configuration
MODEL_NAME = "deepseek-ai/deepseek-r1-6b-chat"
MAX_NEW_TOKENS = 512
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

# Initialize model and tokenizer
try:
    tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
    model = AutoModelForCausalLM.from_pretrained(
        MODEL_NAME,
        device_map="auto",
        torch_dtype=torch.bfloat16 if DEVICE == "cuda" else torch.float32
    )
    print("Model loaded successfully!")
except Exception as e:
    print(f"Model loading failed: {str(e)}")
    model = None

def generate_response(prompt):
    inputs = tokenizer(prompt, return_tensors="pt").to(DEVICE)
    outputs = model.generate(
        **inputs,
        max_new_tokens=MAX_NEW_TOKENS,
        do_sample=True,
        temperature=0.7,
        top_p=0.9,
        pad_token_id=tokenizer.eos_token_id
    )
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

@app.route('/chat', methods=['POST'])
def chat():
    if not model:
        return jsonify({"error": "Model not loaded"}), 500
    
    data = request.json
    prompt = data.get("prompt", "")
    
    if not prompt:
        return jsonify({"error": "No prompt provided"}), 400
    
    try:
        response = generate_response(prompt)
        return jsonify({"response": response})
    
    except Exception as e:
        return jsonify({"error": str(e)}), 500

@app.route('/health', methods=['GET'])
def health_check():
    status = "ready" if model else "unavailable"
    return jsonify({"status": status})

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000)