Skip to content
Test CUDA Builds

AutoTune MoE kernel block sizes for accelerate inference (#18551) #12276

Sign in to view logs

AutoTune MoE kernel block sizes for accelerate inference (#18551)

AutoTune MoE kernel block sizes for accelerate inference (#18551) #12276

Run time

Learn about OS pricing on GitHub Actions
Job Run time
unittest-cuda / linux-job
34m 32s
test-models-cuda / linux-job
47m 34s
test-executorch-cuda-build-12.8 / linux-job
20m 5s
test-executorch-cuda-build-12.9 / linux-job
21m 29s
test-executorch-cuda-build-12.6 / linux-job
19m 45s
test-executorch-cuda-build-13.0 / linux-job
19m 16s
export-model-cuda-artifact (mistralai, Voxtral-Mini-3B-2507, quantized-int4-tile-packed) / linux-job
36m 10s
export-model-cuda-artifact (facebook, dinov2-small-imagenet1k-1-layer, non-quantized) / linux-job
21m 15s
export-model-cuda-artifact (google, gemma-3-4b-it, non-quantized) / linux-job
36m 41s
export-model-cuda-artifact (SocialLocalMobile, Qwen3.5-35B-A3B-HQQ-INT4, quantized-int4-tile-packed) / linux-job
1h 9m 28s
export-model-cuda-artifact (Qwen, Qwen3-0.6B, quantized-int4-tile-packed) / linux-job
25m 15s
export-model-cuda-artifact (mistralai, Voxtral-Mini-3B-2507, quantized-int4-weight-only) / linux-job
34m 15s
export-model-cuda-artifact (mistralai, Voxtral-Mini-3B-2507, non-quantized) / linux-job
34m 1s
export-model-cuda-artifact (nvidia, parakeet-tdt, quantized-int4-tile-packed) / linux-job
31m 29s
export-model-cuda-artifact (openai, whisper-large-v3-turbo, quantized-int4-weight-only) / linux-job
25m 27s
export-model-cuda-artifact (openai, whisper-small, non-quantized) / linux-job
24m 0s
export-model-cuda-artifact (openai, whisper-large-v3-turbo, quantized-int4-tile-packed) / linux-job
26m 14s
export-model-cuda-artifact (nvidia, diar_streaming_sortformer_4spk-v2, non-quantized) / linux-job
35m 50s
export-model-cuda-artifact (Qwen, Qwen3-0.6B, non-quantized) / linux-job
24m 56s
export-model-cuda-artifact (google, gemma-3-4b-it, quantized-int4-tile-packed) / linux-job
35m 30s
export-model-cuda-artifact (nvidia, parakeet-tdt, non-quantized) / linux-job
29m 4s
export-model-cuda-artifact (mistralai, Voxtral-Mini-4B-Realtime-2602, quantized-int4-tile-packed) / linux-job
31m 31s
check-all-cuda-builds
4s
test-cuda-pybind (gemma3-4b, --quantize, google-gemma-3-4b-it-cuda-quantized-int4-tile-packed) / linux-job
24m 42s
test-cuda-pybind (qwen3-0.6b, Qwen-Qwen3-0.6B-cuda-non-quantized) / linux-job
22m 37s
test-cuda-pybind (qwen3-0.6b, --quantize, Qwen-Qwen3-0.6B-cuda-quantized-int4-tile-packed) / linux-job
22m 34s
test-model-cuda-e2e (mistralai, Voxtral-Mini-4B-Realtime-2602, quantized-int4-tile-packed) / linux-job
31m 50s
test-model-cuda-e2e (mistralai, Voxtral-Mini-3B-2507, quantized-int4-tile-packed) / linux-job
17m 12s
test-model-cuda-e2e (openai, whisper-large-v3-turbo, quantized-int4-weight-only) / linux-job
31m 56s
test-model-cuda-e2e (mistralai, Voxtral-Mini-3B-2507, quantized-int4-weight-only) / linux-job
18m 35s
test-model-cuda-e2e (google, gemma-3-4b-it, quantized-int4-tile-packed) / linux-job
17m 27s
test-model-cuda-e2e (nvidia, parakeet-tdt, non-quantized) / linux-job
17m 32s
test-model-cuda-e2e (mistralai, Voxtral-Mini-3B-2507, non-quantized) / linux-job
19m 23s
test-model-cuda-e2e (facebook, dinov2-small-imagenet1k-1-layer, non-quantized) / linux-job
16m 54s
test-model-cuda-e2e (google, gemma-3-4b-it, non-quantized) / linux-job
18m 48s
test-model-cuda-e2e (openai, whisper-small, non-quantized) / linux-job
32m 38s
test-model-cuda-e2e (nvidia, parakeet-tdt, quantized-int4-tile-packed) / linux-job
16m 55s
test-model-cuda-e2e (openai, whisper-large-v3-turbo, quantized-int4-tile-packed) / linux-job
31m 51s
test-model-cuda-e2e (nvidia, diar_streaming_sortformer_4spk-v2, non-quantized) / linux-job
16m 54s
test-model-cuda-e2e (SocialLocalMobile, Qwen3.5-35B-A3B-HQQ-INT4, quantized-int4-tile-packed) / linux-job
39m 55s
18h 1m 34s