DeepSpeed
zero3 performance optimizations
#3622
Merged

zero3 performance optimizations #3622

tjruwase merged 13 commits into deepspeedai:master from BacharL:perf1
BacharL
BacharL Remove dead code
e1e4b753
BacharL Prevent evaluation of debug strings
e3dbb7a3
BacharL BacharL requested a review from jeffra jeffra 2 years ago
BacharL BacharL requested a review from tjruwase tjruwase 2 years ago
BacharL BacharL requested a review from samyam samyam 2 years ago
BacharL BacharL requested a review from mrwyattii mrwyattii 2 years ago
BacharL BacharL force pushed from 3b588087 to f8c7ad3a 2 years ago
BacharL BacharL force pushed from f8c7ad3a to a942dfa1 2 years ago
BacharL BacharL force pushed from a942dfa1 to 3a87dfa2 2 years ago
BacharL Use contiguous gradients tensor reduce scatter between ranks
bd4d724a
BacharL BacharL force pushed from 3a87dfa2 to d6a8711a 2 years ago
BacharL BacharL force pushed from d6a8711a to bd4d724a 2 years ago
BacharL move overflow tracker to optimizer.step
9cf826df
tjruwase
tjruwase commented on 2023-05-30
tjruwase
tjruwase commented on 2023-05-30
tjruwase
tjruwase commented on 2023-05-30
tjruwase Merge branch 'master' into perf1
6847b306
tjruwase
tjruwase commented on 2023-05-30
tjruwase
tjruwase commented on 2023-05-30
tjruwase Merge branch 'master' into perf1
9b80d605
tjruwase
tjruwase Merge branch 'master' into perf1
a813c774
tjruwase Merge branch 'master' into perf1
b0bfdc1a
tjruwase Merge branch 'master' into perf1
7b832241
tjruwase Merge branch 'master' into perf1
89add18d
tjruwase Merge branch 'master' into perf1
16e6f1cf
tjruwase
tjruwase approved these changes on 2023-06-07
tjruwase Merge branch 'master' into perf1
290bdbd9
BacharL Merge branch 'master' into perf1
df88b940
tjruwase tjruwase merged 0977106a into master 2 years ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone