- 26 6月, 2022 1 次提交
-
-
由 Sing_chan 提交于
-
- 15 6月, 2022 1 次提交
-
-
由 Yiqun Liu 提交于
* Optimize prod's python implementation for dygraph. * Change key_dim to head_dim. * Add comment in unittest. * Disable TF32 in unittest.
-
- 08 6月, 2022 1 次提交
-
-
由 Yiqun Liu 提交于
* Polish codes and memory usage for fused_gate_attention. * Fix wrong reduce_dims in fused_gate_attention when computing gradient of nonbatched_bias.
-
- 30 5月, 2022 1 次提交
-
-
由 crystal 提交于
-