Skip to content

Pull requests: ml-explore/mlx

New pull request New
59 Open 2,220 Closed
59 Open 2,220 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Fix gaussian_nll_loss documented default reduction (none -> mean)
#3715 opened Jun 18, 2026 by Pablosinyores Contributor Loading…
fix(metal): end encoding before addCompletedHandler for macOS 26 compatibility
#3714 opened Jun 18, 2026 by vernonstinebaker Loading…
4 tasks done
Improves the MLX-vs-PyTorch benchmark runner for automate…
#3713 opened Jun 18, 2026 by aeiwz Loading…
4 tasks
Fix optimizer epsilon validation message and AdaDelta eps doc
#3712 opened Jun 17, 2026 by Pablosinyores Contributor Loading…
1
Implement jvp for cumulative logsumexp
#3711 opened Jun 17, 2026 by obchain Contributor Loading…
3 of 4 tasks
1
1
Optimize low-channel 3D convolutions via 2D slices
#3709 opened Jun 17, 2026 by chrismicah Contributor Loading…
Fix logsumexp jvp to reduce along the axis
#3708 opened Jun 17, 2026 by obchain Contributor Loading…
3 of 4 tasks
1
[CUDA] JIT-compile qmm_sm80/qmm_sm90/gather_gemm
#3706 opened Jun 17, 2026 by zcbenz Collaborator Loading…
1
Make gemv JIT compilable
#3705 opened Jun 16, 2026 by angeloskath Member Loading…
2
[CUDA] [Improvement] Rope without copy
#3704 opened Jun 16, 2026 by nastya236 Collaborator Loading…
1
[CUDA] Disable managed memory on Tegra
#3701 opened Jun 16, 2026 by gmondada Loading…
2 of 4 tasks
3
Guard scatter_axis against 64-bit outputs on the GPU
#3695 opened Jun 15, 2026 by obchain Contributor Loading…
3 of 4 tasks
1
1
Add more array API functions
#3684 opened Jun 14, 2026 by katlun-lgtm Contributor Loading…
4 tasks done
1
Add flip and unstack
#3683 opened Jun 14, 2026 by katlun-lgtm Contributor Loading…
4 tasks done
8
Add metadata to exported functions
#3682 opened Jun 14, 2026 by Ritesh102000 Loading…
1
1
Add array API standard aliases for existing ops
#3678 opened Jun 14, 2026 by katlun-lgtm Contributor Loading…
4 tasks done
2
Add antialias support to nn.Upsample for linear and cubic modes
#3677 opened Jun 14, 2026 by lyonsno Loading…
Fix state corruption when a primitive throws during eval
#3675 opened Jun 13, 2026 by WindChimeRan Loading…
4 tasks done
2
Fix shapeless compile eliding reductions over size-1 dimensions
#3672 opened Jun 13, 2026 by discobot Contributor Loading…
3 of 4 tasks
1
Optimize CPU masked_scatter for contiguous arrays
#3670 opened Jun 13, 2026 by AK-Khan02 • Draft
1
1
Fused SDPA full-attention for head_dim=192/256 (revival of #3293)
#3660 opened Jun 11, 2026 by hojin12312 Loading…
3 of 4 tasks
1
Fix multi-wire recv prefill deadlock in jaccl ring backend
#3654 opened Jun 10, 2026 by jasonpaulso Loading…
2 of 4 tasks
1
fix(shapeless): infer dynamic dim in Unflatten::output_shapes (#2607)
#3649 opened Jun 9, 2026 by nac7 Loading…
1
2
Fix CPU gather transposing column-contiguous slices
#3647 opened Jun 9, 2026 by tillahoffmann Contributor Loading…
1
Add stft and istft to mlx.core.fft
#3639 opened Jun 8, 2026 by eyupcanakman Contributor Loading…
4 tasks done
6
ProTip! no:milestone will show everything without a milestone.