transformers
100bbcb7 - GPT2 should not store/compute cached activations during finetuning

Commit

6 years ago

GPT2 should not store/compute cached activations during finetuning

Author

thomwolf

thomwolf

Parents

Loading