app.py

from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
import torch
import gradio as gr

# Step 1: Define the model name from Hugging Face Hub
model_name = "deepseek-ai/deepseek-7b-instruct"

# Step 2: Load the tokenizer and model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.float16,  # Use float16 for efficiency
    device_map="auto"           # Automatically assigns GPU or CPU
)

# Step 3: Define a simple function to generate model responses
def chat_function(prompt):
    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
    outputs = model.generate(**inputs, max_length=300, do_sample=True, temperature=0.7)
    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return response

# Step 4: Create the Gradio interface
iface = gr.Interface(
    fn=chat_function,
    inputs=gr.Textbox(lines=5, placeholder="Type your question here..."),
    outputs="text",
    title="🦾 DeepSeek LLM Assistant",
    description="Ask me anything! Powered by DeepSeek-7B-Instruct 🪐"
)

# Step 5: Launch the app
if __name__ == "__main__":
    iface.launch()