- 24 Oct, 2019 1 commit
-
-
Lysandre authored
-
- 22 Oct, 2019 8 commits
-
-
Lysandre Debut authored
-
Lysandre authored
-
Lysandre authored
-
Lysandre authored
-
Juli谩n Peller (dataista) authored
-
Lysandre authored
-
Lysandre authored
-
Thomas Wolf authored
Gradient norm clipping should be done right before calling the optimiser
-
- 21 Oct, 2019 3 commits
-
-
Pasquale Minervini authored
gradient norm clipping should be done right before calling the optimiser - fixing run_glue and run_ner as well
-
thomwolf authored
-
Thomas Wolf authored
Fix hanging when loading pretrained models
-
- 20 Oct, 2019 2 commits
-
-
Pasquale Minervini authored
-
Pasquale Minervini authored
update
-
- 19 Oct, 2019 1 commit
-
-
Ralph Tang authored
- Fix hanging when loading pretrained models from the cache without having internet access. This is a widespread issue on supercomputers whose internal compute nodes are firewalled.
-
- 18 Oct, 2019 3 commits
-
-
LysandreJik authored
-
LysandreJik authored
-
Lysandre Debut authored
-
- 17 Oct, 2019 2 commits
-
-
VictorSanh authored
-
leo-du authored
-
- 16 Oct, 2019 1 commit
-
-
Lysandre Debut authored
Co-authored-by: Thomas Wolf @thomwolf
-
- 15 Oct, 2019 19 commits
-
-
-
thomwolf authored
-
Thomas Wolf authored
Force einsum to run in fp16
-
Thomas Wolf authored
remove leftover usage of DUMMY_INPUTS
-
thomwolf authored
-
Thomas Wolf authored
Added performance enhancements (XLA, AMP) to examples
-
thomwolf authored
-
Thomas Wolf authored
Fixed the sample code in the title 'Quick tour'.
-
Thomas Wolf authored
Remove unnecessary use of FusedLayerNorm in XLNet
-
Thomas Wolf authored
Implement fine-tuning BERT on CoNLL-2003 named entity recognition task
-
thomwolf authored
-
Julien Chaumond authored
-
thomwolf authored
-
thomwolf authored
-
Marianne Stecklina authored
-
Marianne Stecklina authored
-
Marianne Stecklina authored
-
Marianne Stecklina authored
-
Marianne Stecklina authored
-
