microsoft/DeepSpeed

Pull Requests Commits

remove unneeded comments and add zero 2

delock committed 230 days ago

dcf7af9a

Merge branch 'master' into gma/xpu_compile_analysis

delock committed 231 days ago

Verified 12af27d4

Move V100 workflows from cuda 11.1/11.7 to 12.1 (#6607)

loadams committed 232 days ago

Verified 20695b39

Add SSF Best practices badge (#6604)

loadams committed 232 days ago

Verified 940887de

Cleanup CODEOWNERS file to be valid (#6603)

loadams committed 232 days ago

Verified 239b83a7

delock committed 234 days ago

80d0a44f

delock committed 234 days ago

72b633b6

save | literal in table

delock committed 234 days ago

c5e9985e

debug output format

delock committed 234 days ago

c17a2c33

output to git summary file

delock committed 234 days ago

05fa0a85

print summary in markdown table format

delock committed 234 days ago

20022e16

[ROCm] Fix subprocess error (#6587)

jagadish-amd committed 235 days ago

Verified b93c7a20

delock committed 239 days ago

dadc42cd

don't substract

delock committed 239 days ago

b4be3025

avoid grep graph breaks in the workflow

delock committed 239 days ago

eabc834b

no print 'graph_breaks' in dynamo_stats

delock committed 239 days ago

ab3f72e4

delock committed 239 days ago

fe700cfc

print detaild counters

delock committed 239 days ago

3848abc7

force torch dynamo dump more logs

delock committed 240 days ago

5f61c31c

add more logging code in the test

delock committed 240 days ago

45704ae4

print graph log

delock committed 240 days ago

12d19cad

Fix torch include in `op_builder/mlu/fused_adam.py` and update no-torch workflow triggers (#6584)

loadams committed 242 days ago

Verified 8cded575

Fixes on the accelerate side mean we do not need to skip this test (#6583)

loadams committed 242 days ago

Verified 828ddfbb

[COMPILE] workflow for deepspeed + torch.compile (#6570)

YizhouZ committed 243 days ago

Verified d4e18950

add bfloat16 to inference support dtypes (#6528)

nelyahu committed 243 days ago

Verified 1caf6e81

Add APIs to offload states of model, optimizer, and engine (#6011)

tohtana committed 243 days ago

Verified 047bcf6a

[XPU] Support DeepNVMe new code structure (#6532)

Liangliang-Ma committed 243 days ago

Verified d45cfd34

fix errors when setting zero3 leaf modules with torch.compile (#6564)

NirSonnenschein committed 243 days ago

Verified ba58682a

Fix gradient accumulation for Z2+offload (#6550)

tohtana committed 243 days ago

Verified c85c8703

[Accelerator] Cambricon MLU support (#6472)

Andy666G committed 243 days ago

Verified 0fbe96a5

Older