Update README.md

README.md

@@ -24,32 +24,100 @@ This is a fine-tuned version of LLaMA optimized to respond like Rick Sanchez fro
 
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+def setup_rick_model(model_id, use_token=False):
+    """
+    Setup the Rick model from Hugging Face
+    model_id: "username/model-name" from Hugging Face
+    use_token: Set True if it's a private repository
+    """
+    try:
+        # If private repository, first login with token
+        if use_token:
+            from huggingface_hub import login
+            token = "your_token_here"  # Your Hugging Face token
+            login(token)
+
+        # Load model and tokenizer
+        model = AutoModelForCausalLM.from_pretrained(
+            model_id,
+            torch_dtype=torch.float16,
+            device_map="auto"
+        )
+        tokenizer = AutoTokenizer.from_pretrained(model_id)
+        
+        return model, tokenizer
+    
+    except Exception as e:
+        print(f"Error loading model: {str(e)}")
+        return None, None
+
+def ask_rick(question, model, tokenizer, max_length=200):
+    """Ask Rick a question"""
+    # Rick's personality prompt
+    role_play_prompt = (
+        "You are Rick Sanchez, a brilliant mad scientist, "
+        "the smartest man in the universe. Always respond as Rick would—"
+        "sarcastic, genius, and indifferent."
+    )
+    
+    # Format input
+    input_text = f"<s>### Instruction:\n{role_play_prompt}\n\n### Input:\n{question}\n\n### Response:\n"
+    
+    # Generate response
+    inputs = tokenizer(input_text, return_tensors="pt").to(model.device)
+    outputs = model.generate(
+        inputs["input_ids"],
+        max_length=max_length,
+        temperature=0.8,
+        top_p=0.9,
+        do_sample=True,
+        repetition_penalty=1.2
+    )
+    
+    # Decode response
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response.split("### Response:")[-1].strip()
+
+# Usage example
+if __name__ == "__main__":
+    # Replace with your model's repository name
+    MODEL_ID = "CrimsonEyes/rick_sanchez_model"
+    
+    # Load model
+    model, tokenizer = setup_rick_model(MODEL_ID)
+    
+    if model and tokenizer:
+        # Test questions
+        questions = [
+            "What do you think about space travel, Rick?",
+            "Can you explain quantum physics to me?",
+            "What's your opinion on family?"
+        ]
+        
+        for question in questions:
+            print(f"\nQuestion: {question}")
+            response = ask_rick(question, model, tokenizer)
+            print(f"Rick's response: {response}")
+```
+## For a private repository:
 
-# For merged model
-model = AutoModelForCausalLM.from_pretrained("YOUR_USERNAME/rick-llama")
-tokenizer = AutoTokenizer.from_pretrained("YOUR_USERNAME/rick-llama")
-
-# For PEFT/LoRA model
-from peft import PeftModel, PeftConfig
-base_model = AutoModelForCausalLM.from_pretrained("unsloth/Llama-3.2-3B-Instruct")
-model = PeftModel.from_pretrained(base_model, "YOUR_USERNAME/rick-llama")
-tokenizer = AutoTokenizer.from_pretrained("YOUR_USERNAME/rick-llama")
-
-# Format your input
-text = "What do you think about space travel, Rick?"
-
-# Generate response
-inputs = tokenizer(text, return_tensors="pt")
-outputs = model.generate(
-    **inputs,
-    max_length=200,
-    temperature=0.7,
-    top_p=0.9,
-    repetition_penalty=1.2
-)
-response = tokenizer.decode(outputs[0])
 ```
+# First, get your token from https://huggingface.co/settings/tokens
+from huggingface_hub import login
+login("your_token_here")
 
+MODEL_ID = "username/model-name"  # Replace with your model's repository name
+model, tokenizer = setup_rick_model(MODEL_ID, use_token=True)
+```
+
+## Using the model:
+```
+question = "What do you think about space travel, Rick?"
+response = ask_rick(question, model, tokenizer)
+print(f"Rick's response: {response}")
+```
 ## Limitations
 
 - The model may generate responses that are sarcastic or irreverent