app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM
#from huggingface_hub import login

#login(token="hf_VExbFezQQyzOnbpBoRgNxXjiRfMFTGUyj")

my_token="hf_VExbFezQQyzOnbpBoRgNxXjiRfMFTGUyj"

# Load model and tokenizer
model_name = "meta-llama/CodeLlama-7b-hf"
tokenizer = AutoTokenizer.from_pretrained(model_name, token=my_token)
model = AutoModelForCausalLM.from_pretrained(model_name, token=my_token)

# Define the inference function
def generate_code(prompt):
    # Tokenize the input prompt
    inputs = tokenizer(prompt, return_tensors="pt", truncation=True, padding=True)
    
    # Generate code using the model
    outputs = model.generate(inputs["input_ids"], max_length=100, num_return_sequences=1)
    
    # Decode the generated output to a string
    generated_code = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return generated_code

# Create Gradio interface
interface = gr.Interface(
    fn=generate_code, 
    inputs="text", 
    outputs="text",
    title="CodeLlama-7b Python Code Generator",
    description="Generate Python code using the CodeLlama-7b model. Simply input a prompt and get back the generated code.",
)

# Launch the Gradio interface
interface.launch()