deepspeedai/DeepSpeed

Pull Requests Commits

Address tohtana's review feedback for XLA/TPU support

PKUWZP committed 54 days ago

b14e5b46

Address review feedback for XLA/TPU support

PKUWZP committed 54 days ago

7f6ea955

Relax XLA backend test initialization assertion

PKUWZP committed 57 days ago

65cf60fd

Import XLA test target module explicitly

PKUWZP committed 57 days ago

623df9ea

Patch fake torch_xla before import in test

PKUWZP committed 57 days ago

7debfbbb

Fix XLA backend unit test import

PKUWZP committed 58 days ago

41d0f7e9

Fix formatting hooks for XLA backend test

PKUWZP committed 58 days ago

dbfd0a91

Fix TPU device selection for XLA workers

PKUWZP committed 58 days ago

7f82c20d

Add torch_xla TPU support for ZeRO-1/2

PKUWZP committed 59 days ago

8acc916e

Update version (#7903)

loadams committed 67 days ago

Verified 5f7b6870

fix(zero3): use current_stream() instead of default_stream() for grad… (#7898)

michaelroyzen committed 67 days ago

Verified be60451f

Add AGENTS.md and CLAUDE.md with project rules for AI coding agents (#7902)

delock committed 67 days ago

Verified 49a82a0d

fix: Validate fp16.loss_scale is finite and non-negative (#7889)

nathon-lee committed 67 days ago

Verified 63eeb114

Fix Evoformer's multi-arch dispatch root cause (#7881)

tohtana committed 67 days ago

Verified 784cc26e

Fix async_io ops building error on Huawei Ascend NPU (#7894)

huangyifan0610 committed 68 days ago

Verified f88d0f8d

double reduction user-friendly error (#7895)

stas00 committed 70 days ago

Verified b6346bf0

[Bloom] Fix hangs of bloom test (#7890)

k-artem committed 70 days ago

Verified d9a4aadb

Suppress see_memory_usage logs (#7891)

sfc-gh-truwase committed 72 days ago

Verified 285cae30

Fix hook count performance regression from v0.18.5 (#7886)

tohtana committed 75 days ago

Verified 6c59d544

Add document section explaining autocast nesting (#7883)

tohtana committed 76 days ago

Verified 4dba1e20

Remove amp() from abstract accelerator (#7879)

delock committed 78 days ago

Verified a41a96b1

XPU use stock pytorch instead of Intel Extension for PyTorch (#7877)

delock committed 78 days ago

Verified d8e15da4

Fix import deepspeed crash on PyTorch v2.3 + Python 3.12 (#7875)

tohtana committed 78 days ago

Verified 04d69cc6

Fix no-grad grad-fn lookup in ZeRO hook counting on PyTorch 2.3 (#7830) (#7841)

tohtana committed 79 days ago

Verified bffaf457

Fix: only add parameter with grads to parameter group (#7869)

delock committed 79 days ago

Verified 116dbe28

perf: use deque for FIFO queues in sequence parallel, superoffload, and compile (#7880)

giulio-leone committed 79 days ago

Verified a15e557d

fix: replace 35 bare except clauses with except Exception (#7873)

haosenwang1018 committed 82 days ago

Verified ae21699a

Fix broken links and add AutoTP Training tutorial to sidebar nav (#7874)

tohtana committed 82 days ago

Verified efc0b49a

Schedule nightly full test (#7870)

tohtana committed 84 days ago

Verified 0416cf68

Fix Zero-3 static scale assertion in fp16 test (#7866)

tohtana committed 86 days ago

Verified 93524c89

Older