mirrors/icefall
1
0
Fork 0
mirror of https://github.com/k2-fsa/icefall.git synced 2025-09-18 21:44:18 +00:00
Code Issues Packages Projects Releases Wiki Activity

More drafts of new method, not tested.

Browse Source
This commit is contained in:
Daniel Povey 2022-07-06 23:05:06 -07:00
parent 26815d177f
commit e6d00ee3e4
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
egs/librispeech/ASR/pruned_transducer_stateless7/optim.py
View File

@ -715,7 +715,7 @@ class LearnedGradient(Optimizer):
# needed when we rediagonalize which is not that common.
grad_cov.mul_(beta2).add_(torch.matmul(this_g.t(), this_g))
grad = self._project(grad, forward=True, state)
grad = self._project(grad, state, forward=True)
exp_avg_sq = state["exp_avg_sq"]
exp_avg_sq.mul_(beta2).addcmul_(grad, grad,
@ -731,7 +731,7 @@ class LearnedGradient(Optimizer):
grad = grad / denom
# and project back..
grad = self._project(grad, forward=False, state)
grad = self._project(grad, state, forward=False)
scalar_exp_avg_sq = state["scalar_exp_avg_sq"]
@ -782,8 +782,8 @@ class LearnedGradient(Optimizer):
def _project(self,
x: Tensor,
forward: bool,
state: dict) -> Tensor:
state: dict,
forward: bool) -> Tensor:
"""
Multiply a tensor x by proj_{dim} on each of its dimensions.
If forward == True, converts x from canonical to preconditioned/diagonalized