Tri Dao
|
5400fdc4ac
[CE] Implement CrossEntropyLoss in Triton
|
1 year ago |
Tri Dao
|
43ab0b5205
Mention that some CUDA extensions have only been tested on A100s
|
2 years ago |
Tri Dao
|
fa6d1ce44f
Add fused_dense and dropout_add_layernorm CUDA extensions
|
2 years ago |