inference.py

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

# Load base model and tokenizer
base_model = "mistralai/Mistral-7B-v0.1"
tokenizer = AutoTokenizer.from_pretrained(base_model)
model = AutoModelForCausalLM.from_pretrained(
    base_model,
    torch_dtype=torch.float16,
    device_map="auto"
)

# Load LoRA adapter
model = PeftModel.from_pretrained(model, "./")

# Ensure evaluation mode
model.eval()

def chat():
    print("🕉️ Welcome to GodeusAI — your spiritual assistant. Type 'exit' to quit.\n")
    while True:
        user_input = input("You: ")
        if user_input.lower() == "exit":
            print("Goodbye.")
            break

        prompt = f"<|user|>: {user_input}\n<|assistant|>:"
        inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

        with torch.no_grad():
            outputs = model.generate(
                **inputs,
                max_new_tokens=200,
                temperature=0.7,
                top_p=0.9,
                do_sample=True,
                pad_token_id=tokenizer.eos_token_id
            )

        response = tokenizer.decode(outputs[0], skip_special_tokens=True)
        print("GodeusAI:", response.split("<|assistant|>:")[-1].strip())

if __name__ == "__main__":
    chat()