Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jan 21, 2026

• PR 3007 Fix Seg Fault (#3008)

  Show description for 9a277a2
  

  MillaFleurs

  and

  KD2YCU

  authored
  9a277a2

  Copy full SHA for 9a277a2

• [CUDA] Faster grouped mm (#3011)

  zcbenz

  authored
  8017d43

  Copy full SHA for 8017d43

Commits on Jan 20, 2026

• Optimize erf function with expm1f in Metal backend (#3025)

  bjornefisk

  authored
  634b148

  Copy full SHA for 634b148

• Windows CI (#3021)

  zcbenz

  authored
  bfd62a5

  Copy full SHA for bfd62a5

• Fix negative dim indexing (#2994)

  Show description for 83bb789
  

  3 peopleauthored
  83bb789

  Copy full SHA for 83bb789

• Do not give workflow boolean inputs default values (#3014)

  zcbenz

  authored
  65b42c8

  Copy full SHA for 65b42c8

Commits on Jan 19, 2026

• Do not clear disk space in setup-linux (#3013)

  zcbenz

  authored
  0b25c9c

  Copy full SHA for 0b25c9c

• Type Enhancement for Func Transforms and Bug Fix (#3003)

  Show description for 46d0fdc
  

  XXXXRT666

  and

  awni

  authored
  46d0fdc

  Copy full SHA for 46d0fdc

• Fix python package install path in stubgen (#3009)

  zcbenz

  authored
  d96a2bd

  Copy full SHA for d96a2bd

Commits on Jan 18, 2026

• Update CCCL to v3.1.3 (#3012)

  zcbenz

  authored
  9052f67

  Copy full SHA for 9052f67

• Fix sharding of quantized models with non-power-of-2 bits (#3006)

  kernelpool

  authored
  ca14d3d

  Copy full SHA for ca14d3d

Commits on Jan 17, 2026

• fix distributed all_to_sharded bias shard axis from -2 to -1 (#2987)

  gufengc

  authored
  d2bef3c

  Copy full SHA for d2bef3c

Commits on Jan 14, 2026

• Reverts changing the MLX_IBV_DEVICES to MLX_JACCL_DEVICES (#2999)

  angeloskath

  authored
  3fe7794

  Copy full SHA for 3fe7794

• Fix fence (#2998)

  awni

  authored
  4743015

  Copy full SHA for 4743015

• patch bump for next release (#2991)

  awni

  authored
  2469fc2

  Copy full SHA for 2469fc2

Commits on Jan 13, 2026

• Allow some non 2D inputs in qqmm (#2981)

  awni

  authored
  ac26a4c

  Copy full SHA for ac26a4c

• Expose to/from fp8 in Python and don't auto-convert fp8 when loading from safetensors (#2985)

  awni

  authored
  099dcc0

  Copy full SHA for 099dcc0

• Don't try to use NAX at run-time if kernels aren't there (#2982)

  awni

  authored
  8654b82

  Copy full SHA for 8654b82

• Fix RandomBits::is_equivalent to include width (#2978)

  Show description for 4160ec1
  

  4 peopleauthored
  4160ec1

  Copy full SHA for 4160ec1

• replace MLX_IBV_COORDINATOR with MLX_JACCL_COORDINATOR (#2986)

  Evanev7

  authored
  a819779

  Copy full SHA for a819779

Commits on Jan 12, 2026

• fix doc (#2988)

  CC-Yeh

  authored
  7b1c469

  Copy full SHA for 7b1c469

• Add asarray to __array_namespace__ (#2966)

  Anri-Lombard

  authored
  edab937

  Copy full SHA for edab937

• Fix `grid_dim_x` calculations (#2980)

  CC-Yeh

  authored
  46ee0e9

  Copy full SHA for 46ee0e9

Commits on Jan 10, 2026

• Swizzle scales (#2979)

  nastya236

  authored
  43341e8

  Copy full SHA for 43341e8

Commits on Jan 7, 2026

• fix array allocator with user buffer and deleter (#2971)

  andresy

  authored
  1596839

  Copy full SHA for 1596839

Commits on Jan 5, 2026

• QQ linear (#2931)

  nastya236

  authored
  5037317

  Copy full SHA for 5037317

• fix numpy dtype bug (#2960)

  awni

  authored
  1680b6f

  Copy full SHA for 1680b6f

• Fix doc issues in `mlx.nn.init.he_normal` and `mlx.nn.hard_tanh` (#2968)

  Show description for 1df6c2a
  

  Redempt1onzzZZ

  and

  awni

  authored
  1df6c2a

  Copy full SHA for 1df6c2a

Commits on Jan 1, 2026

• BUG FIX - Addition of missing parameter in random::uniform (#2963)

  Show description for 8de9ceb
  

  hwiesmann

  and

  Hartwig Wiesmann

  authored
  8de9ceb

  Copy full SHA for 8de9ceb

Commits on Dec 28, 2025

• refactor: use time.perf_counter for consistent and accurate benchmarking (#2943)

  Satyam12singh

  authored
  d9b950e

  Copy full SHA for d9b950e

Commits on Dec 24, 2025

• Fetch nanobind with cmake (#2949)

  zcbenz

  authored
  26dfe4f

  Copy full SHA for 26dfe4f

• [CUDA] Implement gather_mm_rhs (#2902)

  zcbenz

  authored
  1d21d0e

  Copy full SHA for 1d21d0e

Commits on Dec 23, 2025

• Metal/CPU nvfp4 and mxfp8 (#2946)

  awni

  authored
  1eef1d1

  Copy full SHA for 1eef1d1

• Fixes in mlx.distributed_config (#2947)

  angeloskath

  authored
  9cfda1a

  Copy full SHA for 9cfda1a

• Fix float64 size in data_types.rst (#2948)

  pdevine

  authored
  af2fca5

  Copy full SHA for af2fca5

Pagination

Previous

Next