Cuda out of memory about g-mlp

[jaketae]

Hey @winsapdev, seems like you ran into a CUDA out of memory on Colab. CUDA OOM is normally triggered because the model you're dealing with is too large to be dealt with by the GPU. You can most likely resolve it by declaring a smaller model. In this case, try decreasing d_model or d_ffn. Alternatively, if you have access to a larger GPU, you can try out the model there. Hope this helps!

[jaketae]

Closing this for now, feel free to ping me if you have any questions!