Uploading the logging phase format FindWebEDU dataset. This is a 163 million parameter model because the weights are untied here. 77485ec verified hemantvirmani commited on Jun 6
Uploading the model in Hugging Face format trained on Open Web Text data. It's a 124M parameter model. d1cc0f3 verified hemantvirmani commited on Jun 6
Upload tinygpt_pretrained_weights_tied_openwebtext.pt b8bfbc1 verified hemantvirmani commited on Jun 6
Rename pretraining/checkpoint/tinygpt_pretrained_checkpoint_438k.pt to pretraining/checkpoint/tinygpt_pretrained_checkpoint_untied_fineweb.pt e6ba95f verified hemantvirmani commited on Jun 6