Added changes for analysis Adam.
This commit is contained in:
parent
0c5fa5a637
commit
d2f1672690
|
@ -28,6 +28,7 @@ Features:
|
|||
Features:
|
||||
- Added `skip_zeros` for block-wise and 32-bit optimizers. This ensures correct updates for sparse gradients and sparse models.
|
||||
- Added support for Kepler GPUs. (#4)
|
||||
- Added Analysis Adam to track 8-bit vs 32-bit quantization errors over time.
|
||||
|
||||
Bug fixes:
|
||||
- fixed "undefined symbol: \_\_fatbinwrap_38" error for P100 GPUs on CUDA 10.1 (#5)
|
||||
|
|
Loading…
Reference in New Issue
Block a user