# vx2_comm_gen

This model is a Vision-Language Transformer trained for **cricket commentary generation** from video and token inputs.

## Model Details:
- **Framework**: PyTorch Lightning
- **Vocab Size**: 30522
- **Learning Rate**: 1e-5

## How to Use:
Use your custom `MyTransformerModel` class to load the checkpoint:

    from my_model import MyTransformerModel
    model = MyTransformerModel.load_from_checkpoint("pytorch_model.bin")

## Author
Uploaded by [Mayur Soni](https://huggingface.co/may-ur08)