Abhi Sharma 9e666aaa29 Fix gradient overflow issue during attention mask ...

This fix is in reference to issue #382. GPT2 can now be trained in mixed precision, which I've confirmed with testing. I also tested unconditional generation on multiple seeds before and after changing 1e10 to 1e4 and there was no difference. Please let me know if there is anything else I can do to make this pull request better. Thanks for all your work!

2019-04-16 11:42:34 -07:00

..

__init__.py

added best practices for serialization in README and examples

2019-04-15 15:00:33 +02:00

__main__.py

adding gpt2

2019-02-17 23:38:51 +01:00

convert_gpt2_checkpoint_to_pytorch.py

fix typo - logger info

2019-03-06 10:05:21 +01:00

convert_openai_checkpoint_to_pytorch.py

fix typo - logger info

2019-03-06 10:05:21 +01:00

convert_tf_checkpoint_to_pytorch.py

fix typo - logger info

2019-03-06 10:05:21 +01:00

convert_transfo_xl_checkpoint_to_pytorch.py

fix typo - logger info

2019-03-06 10:05:21 +01:00

file_utils.py

added best practices for serialization in README and examples

2019-04-15 15:00:33 +02:00

modeling_gpt2.py

Fix gradient overflow issue during attention mask

2019-04-16 11:42:34 -07:00

modeling_openai.py

Merge pull request #489 from huggingface/tokenization_serialization

2019-04-16 08:49:54 +02:00

modeling_transfo_xl_utilities.py

fix typo - logger info

2019-03-06 10:05:21 +01:00

modeling_transfo_xl.py

load all models on cpu

2019-04-15 15:43:01 +02:00

modeling.py

Merge pull request #489 from huggingface/tokenization_serialization

2019-04-16 08:49:54 +02:00

optimization_openai.py

same

2019-03-18 15:13:35 +01:00

optimization.py

branches, optim cosine fix

2019-03-18 13:18:07 +01:00

tokenization_gpt2.py

fixing tests

2019-04-15 12:55:38 +02:00

tokenization_openai.py

fix openai special tokens loading

2019-04-15 16:07:45 +02:00

tokenization_transfo_xl.py

tokenization updates

2019-04-15 14:24:52 +02:00

tokenization.py

tokenization updates

2019-04-15 14:24:52 +02:00