app.py

import gradio as gr
import os
from datasets import load_dataset
from huggingface_hub import HfApi, login
from transformers import AutoTokenizer, AutoModelForCausalLM

# Run on NVidia A10G Large (sleep after 1 hour)

# Model IDs:
#
# google/gemma-2-9b-it
# meta-llama/Meta-Llama-3-8B-Instruct

# Datasets:
#
# gretelai/synthetic_text_to_sql

profile = "bstraehle"

def download_model(model_id):
    tokenizer = AutoTokenizer.from_pretrained(model_id)
    model = AutoModelForCausalLM.from_pretrained(model_id)
    model.save_pretrained(model_id)

    return tokenizer

def download_dataset(dataset):
    ds = load_dataset(dataset)
    return ""
    
def fine_tune_model():
    return ""

def upload_model(model_id, tokenizer):
    model_name = model_id[model_id.rfind('/')+1:]
    model_repo_name = f"{profile}/{model_name}"

    login(token=os.environ["HF_TOKEN"])

    api = HfApi()
    api.create_repo(repo_id=model_repo_name)
    api.upload_folder(
        folder_path=model_id,
        repo_id=model_repo_name
    )

    tokenizer.push_to_hub(model_repo_name)

    return model_repo_name

def process(action, system_prompt, user_prompt, model_id, dataset):
    if action == "Prompt base model":
        result = "Prompt base model"
    elif action == "Fine-tuned base model":
        result = "Fine-tuned base model"
    elif action == "Prompt fine-tuned model":
        result = "Prompt fine-tuned model"
    
    #tokenizer = download_model(model_id)
    #model_repo_name = upload_model(model_id, tokenizer)
    
    return result

system_prompt = "You are a text to SQL query translator. Users will ask you a question in English and you will generate a SQL query."
user_prompt = "What is the total trade value and average price for each trader and stock in the trade_history table?"
model_id = "meta-llama/Meta-Llama-3-8B-Instruct"
dataset = "gretelai/synthetic_text_to_sql"

demo = gr.Interface(fn=process, 
                    inputs=[gr.Radio(["Prompt base model", "Fine-tune base model", "Prompt fine-tuned model"], label = "Action")],
                            gr.Textbox(label = "System Prompt", value = system_prompt, lines = 1),
                            gr.Textbox(label = "User Prompt", value = user_prompt, lines = 1),
                            gr.Textbox(label = "Model ID", value = model_id, lines = 1),
                            gr.Textbox(label = "Dataset", value = dataset, lines = 1)],
                    outputs=[gr.Textbox(label = "Completion")])
demo.launch()