Skip to content
periodic

[ET-VK][matmul] Re-implement fp32/fp16 matmul and linear with tiled compute and blocked weight packing #26659

Sign in to view logs

[ET-VK][matmul] Re-implement fp32/fp16 matmul and linear with tiled compute and blocked weight packing

[ET-VK][matmul] Re-implement fp32/fp16 matmul and linear with tiled compute and blocked weight packing #26659

Run time

Learn about OS pricing on GitHub Actions
Job Run time
gather-models
7s
test-models-linux (cmake, mv3, xnnpack-quantization-delegation, linux.2xlarge, 90) / linux-job
11m 26s
test-models-linux (buck2, mv3, portable, linux.2xlarge, 90) / linux-job
11m 10s
test-models-linux (cmake, vit, xnnpack-quantization-delegation, linux.2xlarge, 90) / linux-job
12m 29s
test-models-linux (cmake, mv3, portable, linux.2xlarge, 90) / linux-job
11m 16s
test-models-linux (cmake, vit, portable, linux.2xlarge, 90) / linux-job
18m 1s
test-models-linux (buck2, mv3, xnnpack-quantization-delegation, linux.2xlarge, 90) / linux-job
12m 10s
1h 16m 39s