Tri Dao
|
c7f32a8409
[CrossEntropy] Support precomputed LSE
|
3 bulan lalu |
Curtis "Fjord" Hawthorne
|
d8aacc510c
return z_loss (#768)
|
10 bulan lalu |
Tri Dao
|
713bd3aa9a
[CrossEntropy] Test longer sequences
|
1 tahun lalu |
Tri Dao
|
08124c8f9c
[CrossEntropy] Implement logit_scale option
|
1 tahun lalu |
Tri Dao
|
5400fdc4ac
[CE] Implement CrossEntropyLoss in Triton
|
1 tahun lalu |
Tri Dao
|
0e8c46ae08
Run isort and black on test files
|
1 tahun lalu |
Tri Dao
|
dff68c2b22
Add smoothing for CrossEntropyParallel, rename to CrossEntropyLoss
|
2 tahun lalu |