Skip to content

Actions: pytorch/executorch

Actions

Test XNNPACK Backend

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
test-backend-xnnpack.yml
2,500+ workflow runs
2,500+ workflow runs

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
[ExecuTorch][Llama][ez] Remove unwrap_tensor_subclass from 4w quantization path Test XNNPACK Backend #2569: Pull request #18957 opened by SS-JIA
57m 35s gh/SS-JIA/521/head
gh/SS-JIA/521/head
57m 35s
Qwen 3.5 MoE: Add Metal build target Test XNNPACK Backend #2568: Pull request #18881 synchronize by manuelcandales
1h 0m 19s gh/manuelcandales/176/head
gh/manuelcandales/176/head
1h 0m 19s
Qwen 3.5 MoE: Add --backend metal export path Test XNNPACK Backend #2567: Pull request #18880 synchronize by manuelcandales
1h 0m 5s gh/manuelcandales/175/head
gh/manuelcandales/175/head
1h 0m 5s
Qwen 3.5 MoE: Add Metal build target Test XNNPACK Backend #2566: Pull request #18881 synchronize by manuelcandales
3s gh/manuelcandales/176/head
gh/manuelcandales/176/head
3s
Qwen 3.5 MoE Metal: Use max-sized prefill example for dynamic inputs Test XNNPACK Backend #2565: Pull request #18956 opened by manuelcandales
57m 47s gh/manuelcandales/177/head
gh/manuelcandales/177/head
57m 47s
Add CUDA graph capture/replay for qwen 3.5 moe decode method Test XNNPACK Backend #2564: Pull request #18809 synchronize by Gasoonjia
59m 10s cuda-graph
cuda-graph
59m 10s
Add a persistent Parakeet helper for low-latency host integrations Test XNNPACK Backend #2563: Pull request #18861 synchronize by seyeong-han
58m 24s parakeet-helper-macos-alignment
parakeet-helper-macos-alignment
58m 24s
Arm backend: Handle +FP+INT for vgf and quantized IO Test XNNPACK Backend #2562: Commit 2eff4f4 pushed by per
1h 3m 46s main
main
1h 3m 46s
[ET Device Support] Parse device info from serialized tensor in tensor_parser Test XNNPACK Backend #2561: Pull request #18328 synchronize by Gasoonjia
1h 1m 25s gh/gasoonjia/143/head
gh/gasoonjia/143/head
1h 1m 25s
[ET Device Support] Annotate device attributes of CUDA backend IO tensors cuda device Test XNNPACK Backend #2560: Pull request #18080 synchronize by Gasoonjia
1h 0m 44s gh/gasoonjia/137/head
gh/gasoonjia/137/head
1h 0m 44s
Add Qwen 3.6 MoE model and switch CI to Qwen3.6-35B-A3B-HQQ-INT4 Test XNNPACK Backend #2559: Pull request #18955 opened by mergennachin
1h 1m 34s qwen3_6
qwen3_6
1h 1m 34s
Add per-layer MLP type support for executorch export (#18856) Test XNNPACK Backend #2558: Commit 81bc830 pushed by meta-codesync Bot
1h 0m 31s main
main
1h 0m 31s
temp commit please ignore (#18892) Test XNNPACK Backend #2557: Pull request #18892 synchronize by ryan-monroe
1h 2m 56s ryan-monroe:export-D100856130
ryan-monroe:export-D100856130
1h 2m 56s
Add GPU-side Gumbel-max sampling for CUDA graph compatibility Test XNNPACK Backend #2556: Pull request #18844 synchronize by Gasoonjia
57m 32s cuda-graph-sampling
cuda-graph-sampling
57m 32s
Implement FlashDecoding++ async softmax for split-K SDPA Test XNNPACK Backend #2555: Pull request #18867 synchronize by Gasoonjia
54m 45s gasoonjia/flashdecoding-pp-async-softmax
gasoonjia/flashdecoding-pp-async-softmax
54m 45s
temp commit please ignore (#18892) Test XNNPACK Backend #2554: Pull request #18892 synchronize by ryan-monroe
27m 38s ryan-monroe:export-D100856130
ryan-monroe:export-D100856130
27m 38s
Add FuseConcatPass to eliminate redundant concat ops (#18827) Test XNNPACK Backend #2553: Pull request #18827 synchronize by ryan-monroe
1h 0m 48s ryan-monroe:export-D97667069
ryan-monroe:export-D97667069
1h 0m 48s
Raise layer_norm scalar threshold from 256 to 1024 Test XNNPACK Backend #2552: Pull request #18954 opened by xiaodong705
1h 0m 14s xiaodong705:export-D101050373
xiaodong705:export-D101050373
1h 0m 14s
Interface for swapping inputs and outputs based on cost model + working permute example (#18929) Test XNNPACK Backend #2551: Pull request #18929 synchronize by DrJessop
56m 6s DrJessop:export-D100917820
DrJessop:export-D100917820
56m 6s
Interface for swapping inputs and outputs based on cost model + working permute example (#18929) Test XNNPACK Backend #2550: Pull request #18929 synchronize by DrJessop
15s DrJessop:export-D100917820
DrJessop:export-D100917820
15s
Add MLX op handler for aten.isinf Test XNNPACK Backend #2549: Pull request #18936 synchronize by Ai-chan-0411
1h 1m 12s Ai-chan-0411:fix/mlx-isinf-handler
Ai-chan-0411:fix/mlx-isinf-handler
1h 1m 12s
Fix torchao tests Test XNNPACK Backend #2548: Pull request #18953 opened by metascroy
59m 37s metascroy/fix-torchao-test
metascroy/fix-torchao-test
59m 37s
try to use KA optimize moe kernel Test XNNPACK Backend #2547: Pull request #18930 synchronize by Gasoonjia
1h 6m 6s kamoe
kamoe
1h 6m 6s
Add MLX op handler for aten.isnan Test XNNPACK Backend #2546: Pull request #18952 opened by Jah-yee
1h 2m 15s Jah-yee:add-mlx-op-handler-aten-isnan
Jah-yee:add-mlx-op-handler-aten-isnan
1h 2m 15s
try to use KA optimize moe kernel Test XNNPACK Backend #2545: Pull request #18930 synchronize by Gasoonjia
14m 43s kamoe
kamoe
14m 43s
ProTip! You can narrow down the results and go further in time using created:<2026-04-16 or the other filters available.