app.py

import gradio as gr #web interface
from transformers import AutoModelForCausalLM, AutoTokenizer #for loading the model and making the input into tokens
model_name="Salesforce/codegen-350M-multi"

#initialize the tokenizer and model
tokenizer=AutoTokenizer.from_pretrained(model_name)
model=AutoModelForCausalLM.from_pretrained(model_name)

def generate_code(prompt, max_length=100, temperature=0.7, top_p=0.95):
    inputs=tokenizer(prompt, return_tensors='pt')
    outputs=model.generate(**inputs, max_length=max_length, temperature=temperature, top_p=top_p, do_sample=True) #input: input_id, weight_number

    generated_code=tokenizer.decode(outputs[0],skip_special_tokens=True) #skip_special_tokens will remove <EOS> and <eol>
    return generated_code

#gradio interface
with gr.Blocks() as demo:
    gr.Markdown("## CODE GENERATION WITH CODEGEN MODEL")

    #input box to add prompt
    prompt=gr.Textbox(lines=10, label='enter your prompt for code generation')
    max_length=gr.Slider(50,500, value=100, label='Max Length')
    temperature=gr.Slider(0.1,0.9, value=0.7, label='Temperature')
    top_p=gr.Slider(0.1,1.0, value=0.95,label='Top P value')

    output_box=gr.Textbox(lines=20, label='generated_code')

    generate_button = gr.Button("Generate Code")

    generate_button.click(
        fn=generate_code,
        inputs=[prompt, max_length, temperature, top_p],
        outputs=output_box
    )

demo.launch()