Download Latest Version
v0.31.2 source code.tar.gz (4.3 MB)
Get an email when there's a new version of mlx
| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| README.md | 2026-03-11 | 3.2 kB | |
| v0.31.1 source code.tar.gz | 2026-03-11 | 4.2 MB | |
| v0.31.1 source code.zip | 2026-03-11 | 4.6 MB | |
| Totals: 3 Items | 8.7 MB | 2 | |
What's Changed
- Bump the patch version by @angeloskath in https://github.com/ml-explore/mlx/pull/3185
- Skip Hopper-only kernels in CI by @zcbenz in https://github.com/ml-explore/mlx/pull/3184
- [CUDA] Fsdp (easy) by @nastya236 in https://github.com/ml-explore/mlx/pull/3130
- Fix ref leak in mx.save/load with file like object by @aisk in https://github.com/ml-explore/mlx/pull/3187
- Fix/missing libs in docs by @ChristophePRAT in https://github.com/ml-explore/mlx/pull/3190
- feat: adding the bartlett function by @Vlor999 in https://github.com/ml-explore/mlx/pull/3155
- Bump actions/download-artifact from 7 to 8 by @dependabot[bot] in https://github.com/ml-explore/mlx/pull/3189
- Bump actions/upload-artifact from 6 to 7 by @dependabot[bot] in https://github.com/ml-explore/mlx/pull/3188
- [CUDA] Quantized GEMV by @zcbenz in https://github.com/ml-explore/mlx/pull/3180
- [CUDA] Use fp16 accumulation for 4-bit quant in GEMV by @zcbenz in https://github.com/ml-explore/mlx/pull/3197
- [CUDA] implement Hadamard transform by @Lyxot in https://github.com/ml-explore/mlx/pull/3179
- Improve mlx.distributed_config by @angeloskath in https://github.com/ml-explore/mlx/pull/3199
- PR#3226 Fix by @MillaFleurs in https://github.com/ml-explore/mlx/pull/3227
- PR [#3220] LayerNorm VJP returns zeros_like(weight) instead of zeros_like(bias placeholder) by @MillaFleurs in https://github.com/ml-explore/mlx/pull/3231
- [CUDA] Faster compilation and batch support in QMV by @zcbenz in https://github.com/ml-explore/mlx/pull/3213
- Validate num_splits in split by @MillaFleurs in https://github.com/ml-explore/mlx/pull/3234
- Fix return value in einsum_path for simple contractions by @MillaFleurs in https://github.com/ml-explore/mlx/pull/3232
- Validate dims in rope by @MillaFleurs in https://github.com/ml-explore/mlx/pull/3230
- Fix assigning bool to float16/bfloat16 by @MillaFleurs in https://github.com/ml-explore/mlx/pull/3229
- Fix load_weights with strict=False to filter extra weights before update by @gmin7 in https://github.com/ml-explore/mlx/pull/3214
- Remove custom fp4/fp8 classes by @zcbenz in https://github.com/ml-explore/mlx/pull/3212
- [CUDA] Support 3/5/6-bit quants in QMV by @zcbenz in https://github.com/ml-explore/mlx/pull/3236
- Hybrid sharding by @nastya236 in https://github.com/ml-explore/mlx/pull/3194
- win: fix cuda build by @dhiltgen in https://github.com/ml-explore/mlx/pull/3204
- Remove quantized_utils.cuh by @zcbenz in https://github.com/ml-explore/mlx/pull/3237
- [CUDA] Implement SegmentedMM by @Lyxot in https://github.com/ml-explore/mlx/pull/3238
- Add initial tuning for M5 pro and max by @jagrit06 in https://github.com/ml-explore/mlx/pull/3211
- [CUDA] Use qmv kernel for fp quantizations by @zcbenz in https://github.com/ml-explore/mlx/pull/3239
New Contributors
- @ChristophePRAT made their first contribution in https://github.com/ml-explore/mlx/pull/3190
- @Lyxot made their first contribution in https://github.com/ml-explore/mlx/pull/3179
- @gmin7 made their first contribution in https://github.com/ml-explore/mlx/pull/3214
Full Changelog: https://github.com/ml-explore/mlx/compare/v0.31.0...v0.31.1
