Update app.py

app.py

@@ -1,94 +1,14 @@
-from transformers import AutoModelForCausalLM, AutoTokenizer
-from transformers import BitsAndBytesConfig
-import torch
-import gradio as gr
+RuntimeError: No GPU found. A GPU is needed for quantization.
 
-# Define device
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+During handling of the above exception, another exception occurred:
 
-# Response cache
-response_cache = {
-    "Hi, pretend you are a financial advisor. Now tell me how can I start investing in stock market?": (
-        "As a financial advisor, here’s a guide to start investing in the stock market:\n"
-        "1. **Learn**: Use Investopedia or “The Intelligent Investor” by Benjamin Graham.\n"
-        "2. **Goals**: Set objectives (e.g., retirement) and assess risk tolerance.\n"
-        "3. **Brokerage**: Choose Fidelity (low fees), Vanguard (index funds like VTI), or Robinhood (commission-free).\n"
-        "4. **Investments**: Start with ETFs (e.g., VOO for S&P 500) or mutual funds.\n"
-        "5. **Strategy**: Use dollar-cost averaging with $100-$500 monthly.\n"
-        "6. **Risks**: Diversify and monitor.\n"
-        "Consult a certified financial planner."
-    ),
-    "do you have a list of companies you recommend?": (
-        "I cannot recommend specific companies without current market data. Instead, consider ETFs like VOO (S&P 500) or QQQ (tech-focused) for broad exposure. "
-        "For individual stocks, research sectors like technology (e.g., Apple, Microsoft) or consumer goods (e.g., Procter & Gamble) using Yahoo Finance or Morningstar. "
-        "Always consult a certified financial planner."
-    )
-}
-
-# Load model
-model_name = "facebook/opt-1.3B"
-try:
-    tokenizer = AutoTokenizer.from_pretrained(model_name, clean_up_tokenization_spaces=False)
-    model = AutoModelForCausalLM.from_pretrained(
-        model_name,
-        device_map="auto",
-        torch_dtype=torch.float16,
-        quantization_config=BitsAndBytesConfig(load_in_8bit=True)
-    ).to(device)
-except Exception as e:
-    model_name = "facebook/opt-350m"
-    tokenizer = AutoTokenizer.from_pretrained(model_name, clean_up_tokenization_spaces=False)
+Traceback (most recent call last):
+  File "/home/user/app/app.py", line 41, in <module>
     model = AutoModelForCausalLM.from_pretrained(
-        model_name,
-        device_map="auto",
-        torch_dtype=torch.float16,
-        quantization_config=BitsAndBytesConfig(load_in_8bit=True)
-    ).to(device)
-
-# Define chat function
-def chat_with_model(message, history):
-    try:
-        if not isinstance(message, str):
-            return "Error: User input must be a string"
-        if message in response_cache:
-            return response_cache[message]
-        full_prompt = (
-            "You are a financial advisor with expertise in stock market investments. "
-            "Provide accurate, detailed, and actionable advice. "
-            "If you cannot provide specific recommendations (e.g., individual companies), "
-            "explain why and offer general guidance or alternative suggestions instead.\n"
-        )
-        history = history[-3:] if history else []  # Limit history
-        for user_msg, assistant_msg in history or []:
-            if user_msg:
-                full_prompt += f"User: {user_msg}\n"
-            if assistant_msg:
-                full_prompt += f"Assistant: {assistant_msg}\n"
-        full_prompt += f"User: {message}\nAssistant:"
-        inputs = tokenizer(full_prompt, return_tensors="pt", truncation=True, max_length=512).to(device)
-        outputs = model.generate(
-            **inputs,
-            max_new_tokens=150,
-            do_sample=True,
-            top_p=0.9,
-            temperature=0.5,
-            pad_token_id=tokenizer.eos_token_id
-        )
-        response = tokenizer.decode(outputs[0], skip_special_tokens=True)
-        return response[len(full_prompt):].strip() if response.startswith(full_prompt) else response
-    except Exception as e:
-        return f"Error generating response: {str(e)}"
-
-# Create Gradio interface
-interface = gr.ChatInterface(
-    fn=chat_with_model,
-    title="Financial Advisor Chatbot (OPT-1.3B)",
-    description="Ask for advice on starting to invest in the stock market! Powered by Meta AI's OPT-1.3B.",
-    examples=[
-        "Hi, pretend you are a financial advisor. Now tell me how can I start investing in stock market?",
-        "Do you have a list of companies you recommend?"
-    ]
-)
-
-# Launch interface
-interface.launch()
+  File "/usr/local/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py", line 564, in from_pretrained
+    return model_class.from_pretrained(
+  File "/usr/local/lib/python3.10/site-packages/transformers/modeling_utils.py", line 3398, in from_pretrained
+    hf_quantizer.validate_environment(
+  File "/usr/local/lib/python3.10/site-packages/transformers/quantizers/quantizer_bnb_8bit.py", line 62, in validate_environment
+    raise RuntimeError("No GPU found. A GPU is needed for quantization.")
+RuntimeError: No GPU found. A GPU is needed for quantization.