from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
import os

# Set a writable cache directory for Hugging Face
os.environ['TRANSFORMERS_CACHE'] = '/cache'

# Load LLaMA 3 model and tokenizer
# model_name = "https://huggingface.co/nvidia/Llama3-ChatQA-2-70B"  # Replace with LLaMA 3 model path or Hugging Face model link if available
model_name = "nvidia/Llama3-ChatQA-2-70B"  # Replace with LLaMA 3 model path or Hugging Face model link if available
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

# Define API for text generation
def generate_text(prompt, max_length=100):
    inputs = tokenizer(prompt, return_tensors="pt")
    outputs = model.generate(**inputs, max_length=max_length)
    return tokenizer.decode(outputs[0], skip_special_tokens=True)