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


max_seq_length = 4096 # Choose any! We auto support RoPE Scaling internally!
dtype = None # None for auto detection. Float16 for Tesla T4, V100, Bfloat16 for Ampere+
load_in_4bit = True # Use 4bit quantization to reduce memory usage. Can be False.

# 4bit pre quantized models we support for 4x faster downloading + no OOMs.
fourbit_models = [
    "unsloth/llama-3-8b-Instruct-bnb-4bit",
            ] 

model, tokenizer = FastLanguageModel.from_pretrained(
    model_name = "unsloth/llama-3-8b-Instruct-bnb-4bit",
    max_seq_length = max_seq_length,
    dtype = dtype,
    load_in_4bit = load_in_4bit,
    # token = "hf_...", # use one if using gated models like meta-llama/Llama-2-7b-hf
)


# Load the base model and apply LoRA adapters
from transformers import AutoModel
adapter_model = AutoModel.from_pretrained("Rohan5manza/sentiment_analysis")

model = PeftModel.from_pretrained(model, adapter_model)

def generate_response(prompt):
    inputs = tokenizer(prompt, return_tensors="pt")
    outputs = model.generate(**inputs)
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

# Example Gradio or Streamlit interface for deploying
import gradio as gr

def gradio_interface(prompt):
    response = generate_response(prompt)
    return response

iface = gr.Interface(fn=gradio_interface, inputs="text", outputs="text")
iface.launch()