DeepSpeed
Optimize grad_norm calculations by reducing device/host dependency
#4974
Merged

Optimize grad_norm calculations by reducing device/host dependency #4974

nelyahu
nelyahu nelyahu requested a review from tjruwase tjruwase 1 year ago
nelyahu nelyahu requested a review from mrwyattii mrwyattii 1 year ago
tjruwase
tjruwase approved these changes on 2024-01-18
nelyahu nelyahu force pushed from 2d828e0d to 7a9229bf 1 year ago
nelyahu nelyahu force pushed from cedf74f2 to 9b424a25 1 year ago
Optimize grad_norm calculations by reducing device/host dependency
6bd8edd4
nelyahu nelyahu force pushed from 727e2c60 to 6bd8edd4 1 year ago
nelyahu Merge branch 'microsoft:master' into stage_1_and_2_perf_opt
94566598
nelyahu
tjruwase Merge branch 'master' into stage_1_and_2_perf_opt
fbd955cb
nelyahu Merge branch 'master' into stage_1_and_2_perf_opt
e5077daf
tjruwase Merge branch 'master' into stage_1_and_2_perf_opt
1dbb7120
mrwyattii Merge branch 'master' into stage_1_and_2_perf_opt
82b6656f
mrwyattii mrwyattii merged 61daaa1e into master 1 year ago
nelyahu nelyahu deleted the stage_1_and_2_perf_opt branch 1 year ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone