Name | Modified | Size | Downloads / Week |
---|---|---|---|
Parent folder | |||
README.md | 2025-05-22 | 1.3 kB | |
v0.16.9 Patch Release source code.tar.gz | 2025-05-22 | 199.2 MB | |
v0.16.9 Patch Release source code.zip | 2025-05-22 | 200.3 MB | |
Totals: 3 Items | 399.6 MB | 1 |
What's Changed
- Update patch version after 0.16.8 release by @loadams in https://github.com/deepspeedai/DeepSpeed/pull/7296
- Avoid graph break by removing another redundant requires grad false by @deepcharm in https://github.com/deepspeedai/DeepSpeed/pull/7263
- Add qwen3 meta loading for AutoTP by @delock in https://github.com/deepspeedai/DeepSpeed/pull/7293
- Modernize system executable detection across components by @emmanuel-ferdman in https://github.com/deepspeedai/DeepSpeed/pull/7290
- Enable ZeRO set/get APIs for NVMe offload by @tjruwase in https://github.com/deepspeedai/DeepSpeed/pull/7046
- Add qwen3moe meta loading for AutoTP by @ranzhejiang in https://github.com/deepspeedai/DeepSpeed/pull/7297
- disable license check until the new license situation has been sorted… by @stas00 in https://github.com/deepspeedai/DeepSpeed/pull/7301
- Fix extra_repr_str when weight is None / in zero-3 by @HollowMan6 in https://github.com/deepspeedai/DeepSpeed/pull/7254
- [XPU] Support XCCL on deepspeed side by @ys950902 in https://github.com/deepspeedai/DeepSpeed/pull/7299
New Contributors
- @emmanuel-ferdman made their first contribution in https://github.com/deepspeedai/DeepSpeed/pull/7290
Full Changelog: https://github.com/deepspeedai/DeepSpeed/compare/v0.16.8...v0.16.9