Update app.py

app.py

@@ -7,26 +7,42 @@ import os
 model_path = "rajj0/autotrain-phi3-midium-4k-godsent-orpo-6"
 hf_token = os.getenv("HF_TOKEN")  # Get the token from environment variables
 
+# Debugging: print the token to ensure it's being set
+print(f"HF_TOKEN: {hf_token}")
+
 if hf_token is None:
     raise ValueError("HF_TOKEN environment variable not set")
 
-# Load the tokenizer and model with trust_remote_code=True
-tokenizer = AutoTokenizer.from_pretrained(model_path, use_auth_token=hf_token, trust_remote_code=True)
-model = AutoModelForCausalLM.from_pretrained(
-    model_path,
-    device_map="auto",
-    torch_dtype='auto',
-    use_auth_token=hf_token,
-    trust_remote_code=True
-).eval()
+try:
+    # Load the tokenizer and model with trust_remote_code=True
+    print("Loading tokenizer...")
+    tokenizer = AutoTokenizer.from_pretrained(model_path, use_auth_token=hf_token, trust_remote_code=True)
+    print("Tokenizer loaded successfully.")
+
+    print("Loading model...")
+    model = AutoModelForCausalLM.from_pretrained(
+        model_path,
+        device_map="auto",
+        torch_dtype='auto',
+        use_auth_token=hf_token,
+        trust_remote_code=True
+    ).eval()
+    print("Model loaded successfully.")
+except Exception as e:
+    print(f"Error loading model or tokenizer: {e}")
+    raise
 
 # Function to generate a response from the model
 def generate_response(user_input):
-    messages = [{"role": "user", "content": user_input}]
-    input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
-    output_ids = model.generate(input_ids.to('cuda'))
-    response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)
-    return response
+    try:
+        messages = [{"role": "user", "content": user_input}]
+        input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
+        output_ids = model.generate(input_ids.to('cuda'))
+        response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)
+        return response
+    except Exception as e:
+        print(f"Error generating response: {e}")
+        return "An error occurred while generating the response."
 
 # Create the Gradio interface
 iface = gr.Interface(
@@ -39,4 +55,6 @@ iface = gr.Interface(
 
 # Launch the Gradio interface
 if __name__ == "__main__":
-    iface.launch()
+    print("Launching Gradio interface...")
+    iface.launch()
+    print("Gradio interface launched.")