app.py

import gradio as gr
from transformers import AutoModelForSeq2SeqLM
from transformers import DataCollatorForSeq2Seq, AutoConfig

model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
print(f"Successfully loaded the model without gradio or spaces, model object: {model}")

@spaces.GPU(duration=120)
def run_train(model_name, dataset_name, hub_id, api_key, num_epochs, batch_size, lr, grad):

    model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
    return "WORKS"
# Create Gradio interface
try:    
    iface = gr.Interface(
        fn=run_train,
        inputs=[
            gr.Textbox(label="Model Name (e.g., 'google/t5-efficient-tiny-nh8')"),
            gr.Textbox(label="Dataset Name (e.g., 'imdb')"),
            gr.Textbox(label="HF hub to push to after training"),
            gr.Textbox(label="HF API token"),
            gr.Slider(minimum=1, maximum=10, value=3, label="Number of Epochs", step=1),
            gr.Slider(minimum=1, maximum=2000, value=1, label="Batch Size", step=1),
            gr.Slider(minimum=1, maximum=1000, value=1, label="Learning Rate (e-5)", step=1),
            gr.Slider(minimum=1, maximum=100, value=1, label="Gradient accumulation", step=1), 
        ],
        outputs="text",
        title="Fine-Tune Hugging Face Model",
        description="This interface allows you to fine-tune a Hugging Face model on a specified dataset."
    )

    # Launch the interface
    iface.launch()    
except Exception as e:
    print(f"An error occurred: {str(e)}, TB: {traceback.format_exc()}")