Tri Dao
|
88173a1aaf
[FusedDense] Support relu, rename FusedDenseGeluDense -> FusedMLP
|
1 gadu atpakaļ |
Tri Dao
|
43798966cf
[Docs] Fix formatting
|
1 gadu atpakaļ |
Tri Dao
|
3c7cbfc195
[Docs] Mention that dropout_layer_norm supports all dims up to 6k
|
1 gadu atpakaļ |
Tri Dao
|
4a6eaa9f27
Update configs, add results
|
2 gadi atpakaļ |
Tri Dao
|
0bf5e50038
Release training code
|
2 gadi atpakaļ |