DeepSpeed
[zero] revert PR #3166, it disabled grad clip for bf16
#3790
Merged

[zero] revert PR #3166, it disabled grad clip for bf16 #3790

tjruwase merged 29 commits into master from revert-3166
jeffra
HeyangQin zero++ tutorial PR (#3783)
df1859d6
zhiruiluo [Fix] _conv_flops_compute when padding is a str and stride=1 (#3169)
d81a6ad6
cli99 fix interpolate flops compute (#3782)
a8c182a4
CaffreyR use `Flops Profiler` to test `model.generate()` (#2515)
c4c442f0
jeffra revert PR #3166, it disabled grad clip for bf16
9bd7b24c
jeffra ensure no loss scaling for non-fp16 dtypes
6075a29d
jeffra jeffra requested a review from tjruwase tjruwase 2 years ago
jeffra jeffra requested a review from samyam samyam 2 years ago
jeffra jeffra requested a review from mrwyattii mrwyattii 2 years ago
jeffra revert PR #3611 (#3786)
fc9e1ee0
jeffra bump to 0.9.6
40045dc7
jeffra Merge branch 'master' into revert-3166
710a59c6
HeyangQin ZeRO++ chinese blog (#3793)
49a0a1bb
jeffra remove staging trigger (#3792)
2c62cb4c
stephen-youn DeepSpeed-Triton for Inference (#3748)
4dc65f7b
HeyangQin ZeRO++ (#3784)
e1119d8b
HeyangQin adding zero++ to navigation panel of deepspeed.ai (#3796)
01b843aa
tohtana Add ZeRO++ Japanese blog (#3797)
319b64ed
cli99 Bug Fixes for autotuner and flops profiler (#1880)
b4a2c0af
cmikeh2 Missing strided copy for gated MLP (#3788)
b7e1010b
jomayeri Requires grad checking. (#3789)
e5b1eadb
jeffra bump to 0.10.0
9c756cf8
rraminen Fix Bug in transform.cu (#3534)
a204edc7
stephen-youn bug fix: triton importing error (#3799)
f6e2e38b
jeffra Merge branch 'master' into revert-3166
5c8bae02
jeffra jeffra force-pushed the master branch from f6e2e38b to bafaf3c0 2 years ago
jeffra jeffra requested a review from awan-10 awan-10 2 years ago
jeffra jeffra requested a review from RezaYazdaniAminabadi RezaYazdaniAminabadi 2 years ago
jeffra jeffra requested a review from cmikeh2 cmikeh2 2 years ago
jeffra jeffra requested a review from arashb arashb 2 years ago
jeffra jeffra requested a review from cli99 cli99 2 years ago
jeffra jeffra requested a review from loadams loadams 2 years ago
jeffra Merge branch 'master' into revert-3166
928dc2c6
guoyejun
tjruwase
tjruwase approved these changes on 2023-06-26
tjruwase Merge branch 'master' into revert-3166
c290d4c1
loadams
loadams approved these changes on 2023-06-26
loadams Merge branch 'master' into revert-3166
25e083a1
tjruwase Merge branch 'master' into revert-3166
cafd8186
tjruwase Merge branch 'master' into revert-3166
f3c44ccc
tjruwase Merge branch 'master' into revert-3166
4854b5c0
tjruwase Merge branch 'master' into revert-3166
a8ffc37f
tjruwase tjruwase merged 691d246e into master 2 years ago
mrwyattii mrwyattii deleted the revert-3166 branch 2 years ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone