Update app.py

app.py

@@ -1,46 +1,62 @@
+import os
+import torch
 import gradio as gr
-from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
+from transformers import AutoTokenizer, AutoModelForCausalLM
 from peft import PeftModel
+from huggingface_hub import login
 
-# Load base + LoRA model
-base_model = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
-lora_model = "Futuresony/future_12_10_2024"
+# Authenticate with Hugging Face using secret HF_TOKEN
+hf_token = os.environ.get("HF_TOKEN")
+if not hf_token:
+    raise RuntimeError("Missing HF_TOKEN in secrets. Please add it in your Space settings.")
 
-tokenizer = AutoTokenizer.from_pretrained(base_model)
-base = AutoModelForCausalLM.from_pretrained(base_model)
-model = PeftModel.from_pretrained(base, lora_model)
+login(token=hf_token)
+
+# Load base model and LoRA adapter
+base_model_id = "unsloth/gemma-2-9b"  # Or your base model
+lora_model_id = "Futuresony/future_12_10_2024"        # Your LoRA fine-tuned model
+
+# Load tokenizer and model
+tokenizer = AutoTokenizer.from_pretrained(base_model_id)
+base_model = AutoModelForCausalLM.from_pretrained(base_model_id, torch_dtype=torch.float16, device_map="auto")
+model = PeftModel.from_pretrained(base_model, lora_model_id)
+
+# Ensure model is in evaluation mode
 model.eval()
 
-# Create generation pipeline
-generator = pipeline("text-generation", model=model, tokenizer=tokenizer)
+def generate_response(message, history, system_message, max_tokens, temperature, top_p):
+    prompt = system_message + "\n\n"
+
+    for user_input, bot_response in history:
+        prompt += f"User: {user_input}\nAssistant: {bot_response}\n"
 
-# Define the chat function
-def respond(message, history, system_message, max_tokens, temperature, top_p):
-    prompt = system_message + "\n"
-    for user, bot in history:
-        prompt += f"User: {user}\nAssistant: {bot}\n"
     prompt += f"User: {message}\nAssistant:"
 
-    response = generator(
-        prompt,
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    outputs = model.generate(
+        **inputs,
         max_new_tokens=max_tokens,
         temperature=temperature,
         top_p=top_p,
         do_sample=True,
-        return_full_text=False,
-    )[0]["generated_text"]
+        pad_token_id=tokenizer.eos_token_id
+    )
 
-    yield response.strip()
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    final_response = response.split("Assistant:")[-1].strip()
+    return final_response
 
-# Set up Gradio UI
+# Gradio ChatInterface
 demo = gr.ChatInterface(
-    respond,
+    fn=generate_response,
     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)"),
+        gr.Textbox(value="You are a helpful assistant.", label="System Message"),
+        gr.Slider(50, 1024, value=256, step=1, label="Max Tokens"),
+        gr.Slider(0.1, 1.5, value=0.7, step=0.1, label="Temperature"),
+        gr.Slider(0.1, 1.0, value=0.95, step=0.05, label="Top-p"),
     ],
+    title="LoRA AI Chat Assistant",
+    description="Chat with your fine-tuned model using LoRA adapter."
 )
 
 if __name__ == "__main__":