deepspeedai/DeepSpeed

Pull Requests Commits

Merge branch 'master' into loadams/cuda-compilation-nv-bfloat162

loadams committed 1 year ago

Verified 134399e2

Update CPU torch version to 2.7 (#7241)

loadams committed 1 year ago

Verified 9926879b

Revert "Update torch cpu test version"

loadams committed 1 year ago

8d2865e0

Update torch cpu test version

loadams committed 1 year ago

00b5678b

Merge branch 'master' into loadams/cuda-compilation-nv-bfloat162

loadams committed 1 year ago

Verified f2b560b5

Add cpu accelerator fp16 dtype support (#7207)

Yejing-Lai committed 1 year ago

Verified d79bd930

[NFC] Fix comment related to SP group (#7234)

c8ef committed 1 year ago

Verified 9c9d32c2

Recommend using latest (#7233)

tohtana committed 1 year ago

Verified 962a8f0a

Update version.txt after 0.16.7 release (#7232)

loadams committed 1 year ago

Verified ff231af7

Make sure it's not None before offloading contiguous_grad_buffer (#7227)

HollowMan6 committed 1 year ago

Verified c66fdaf3

Pass `with_cuda` arg for jit_load in OpBuilder (#7226)

HollowMan6 committed 1 year ago

Verified 625ea528

Add defence for DeepCompile w/o optimizer (#7225)

HollowMan6 committed 1 year ago

Verified 86e51e69

Fix build on AMD GPUs (related to DeepCompile) (#7224)

HollowMan6 committed 1 year ago

Verified 0f224d79

Fix pass for z3 and profiler (#7222)

tohtana committed 1 year ago

Verified 1d022e05

Fix release links (#7219)

tjruwase committed 1 year ago

Verified 8f93f8b9

Update version.txt after 0.16.6 release (#7218)

loadams committed 1 year ago

Verified a01a2688

DeepCompile for enhanced compiler integration (#7154)

tohtana committed 1 year ago

Verified 227a60c0

Merge branch 'master' into loadams/cuda-compilation-nv-bfloat162

hwchen2017 committed 1 year ago

Verified 989fb619

Add defence for offload_states and reload_states w/o optimizer (#7211)

HollowMan6 committed 1 year ago

Verified a21e5b9d

Support complicated use cases with TiedLayerSpec (#7208)

limjcst committed 1 year ago

Verified 185330cd

HPU accelerator memory mapping is broken because of torch fill uninit memory (#7209)

oelayan7 committed 1 year ago

Verified 56005d2b

update dependencies version info (#7206)

inkcherry committed 1 year ago

Verified 29fa95a8

Merge branch 'master' into loadams/cuda-compilation-nv-bfloat162

loadams committed 1 year ago

Verified f66296dc

Update to fix pydantic warning (#7193)

loadams committed 1 year ago

Verified 027ee21a

Reland perf fix for nan inf check (#7184)

nelyahu committed 1 year ago

Verified 3c1817f3

Update to new torch grad hook API: BF16Optimizer and Stage2 (#7189)

deepcharm committed 1 year ago

Verified 79ff1627

Add qwen3 autotp support (#7187)

Yejing-Lai committed 1 year ago

Verified 11aca297

Remove definition that causes error in cuda 12.6

loadams committed 1 year ago

fbaacf35

Fix issue #5242 grad_norm and loss is nan (#7171)

Glaceon-Hyy committed 1 year ago

Verified 1f706621

async tp allreduce (#7115)

inkcherry committed 1 year ago

Verified b8cc1eb0

Older