Skip to content

Actions: ml-explore/mlx

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
2,500+ workflow runs
2,500+ workflow runs

Filter by Workflow

⌥ + click/return to exclude

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Metal 4.1 fixes Build and Test #3028: Pull request #3963 opened by louen
Action required louen:val/fix-metal41
louen:val/fix-metal41
Action required
Fix filter_and_map docstring argument order for filter_fn and is_leaf_fn Build and Test #3027: Pull request #3906 synchronize by ayaangazali
Action required ayaangazali:fix-filter-and-map-doc
ayaangazali:fix-filter-and-map-doc
Action required
Fix InstanceNorm Shape docstring to require at least 3 dimensions Build and Test #3026: Pull request #3903 synchronize by ayaangazali
Action required ayaangazali:fix-instancenorm-shape-doc
ayaangazali:fix-instancenorm-shape-doc
Action required
Fix step activation docstring to match >= threshold behavior Build and Test #3025: Pull request #3902 synchronize by ayaangazali
Action required ayaangazali:fix-step-docstring
ayaangazali:fix-step-docstring
Action required
Fix Transformer ignoring a custom encoder or decoder with no parameters Build and Test #3024: Pull request #3962 opened by ayaangazali
Action required ayaangazali:fix-transformer-custom-modules
ayaangazali:fix-transformer-custom-modules
Action required
Fix mx.random.randint precision loss beyond 2^24 (float32 domain sampling) Build and Test #3023: Pull request #3936 synchronize by PhilipJohnBasile
Action required PhilipJohnBasile:agent/fix-randint-float32-range
PhilipJohnBasile:agent/fix-randint-float32-range
Action required
Fix mx.random.randint precision loss beyond 2^24 (float32 domain sampling) Build and Test #3022: Pull request #3936 synchronize by PhilipJohnBasile
Action required PhilipJohnBasile:agent/fix-randint-float32-range
PhilipJohnBasile:agent/fix-randint-float32-range
Action required
Optimize large NVFP4 QMV on M5 Max Build and Test #3021: Pull request #3961 opened by dhiltgen
Action required dhiltgen:qmv-nvfp4
dhiltgen:qmv-nvfp4
Action required
Fix sorted gather_mm activation row stride Build and Test #3020: Pull request #3960 opened by metascroy
Action required metascroy:fix-gather-mm-rhs-lda
metascroy:fix-gather-mm-rhs-lda
Action required
docs: remove references to removed --no-verify-script launch flag Build and Test #3019: Pull request #3959 opened by latent-9
Action required latent-9:fix-stale-no-verify-script-doc
latent-9:fix-stale-no-verify-script-doc
Action required
Canonicalize the path returned by current_binary_dir() Build and Test #3018: Pull request #3958 opened by GoodOlClint
-1s GoodOlClint:pr/canonicalize-binary-dir
GoodOlClint:pr/canonicalize-binary-dir
-1s
Pick gather_qmm_rhs tile size from rows per expert Build and Test #3017: Pull request #3918 synchronize by spokvulcan
Action required spokvulcan:perf/gather-qmm-rpe-tiles
spokvulcan:perf/gather-qmm-rpe-tiles
Action required
Fuse the causal mask into the SDPA fallback softmax Build and Test #3016: Pull request #3919 synchronize by spokvulcan
Action required spokvulcan:perf/fused-causal-softmax-sdpa-fallback
spokvulcan:perf/fused-causal-softmax-sdpa-fallback
Action required
Faster eval walk and cached gather_mm identity indices Build and Test #3015: Pull request #3920 synchronize by spokvulcan
Action required spokvulcan:perf/eval-hot-path-overhead
spokvulcan:perf/eval-hot-path-overhead
Action required
Faster eval walk and cached gather_mm identity indices Build and Test #3014: Pull request #3920 synchronize by spokvulcan
Action required spokvulcan:perf/eval-hot-path-overhead
spokvulcan:perf/eval-hot-path-overhead
Action required
Fix incorrect examples and outputs in the usage docs (#3956) Build and Test #3013: Commit fb5133e pushed by angeloskath
34m 8s main
main
34m 8s
Implement batched matmul for large 1D dot products Build and Test #3012: Pull request #3580 synchronize by Ved235
Action required Ved235:main
Ved235:main
Action required
Fix incorrect examples and outputs in the usage docs Build and Test #3011: Pull request #3956 opened by ayaangazali
33m 23s ayaangazali:fix-usage-docs-examples
ayaangazali:fix-usage-docs-examples
33m 23s
Sample randint directly instead of through a float32 uniform Build and Test #3010: Pull request #3955 opened by axiom-of-choice
Action required axiom-of-choice:fix/randint-exact-integer-range
axiom-of-choice:fix/randint-exact-integer-range
Action required
[Metal] Skip empty NAX GEMM output groups Build and Test #3009: Pull request #3941 synchronize by XXXXRT666
Action required XXXXRT666:metal-nax-skip-empty
XXXXRT666:metal-nax-skip-empty
Action required
added eye(0) support and tests Build and Test #3008: Pull request #3952 opened by aaishwarymishra
Action required aaishwarymishra:eye
aaishwarymishra:eye
Action required
Release build Release build #191: Scheduled
30m 39s main
main
30m 39s
Add reflect and symmetric padding modes to mx.pad Build and Test #3007: Pull request #3608 synchronize by katlun-lgtm
Action required katlun-lgtm:add-reflect-symmetric-pad
katlun-lgtm:add-reflect-symmetric-pad
Action required
Raise a clear error for an invalid quantization mode in nn layers (#3… Build and Test #3006: Commit 2ad0d4d pushed by nastya236
35m 16s main
main
35m 16s
Fix cholesky_inv documented argument name to match the binding Build and Test #3004: Pull request #3950 opened by ayaangazali
Action required ayaangazali:fix-cholesky-inv-argname
ayaangazali:fix-cholesky-inv-argname
Action required