CI (self-hosted)

fix: allow mixed turbo/q8_0 KV cache types in CUDA flash-attention selection #12

Sign in to view logs

Run time

Learn about OS pricing on GitHub Actions

Job	Run time
gpu-vulkan-nvidia-cm2	1d 0h 0m 0s
gpu-metal	1d 0h 0m 1s
gpu-webgpu-nvidia	1d 0h 0m 1s
gpu-webgpu-apple	1d 0h 0m 1s
gpu-cuda	1d 0h 0m 0s
gpu-vulkan-nvidia-cm	1d 0h 0m 1s
gpu-vulkan-apple	1d 0h 0m 0s
gpu-openvino-low-perf	1d 0h 0m 1s
cpu-arm64-high-perf-graviton4	1d 0h 0m 1s
gpu-vulkan-intel-linux	1d 0h 0m 1s
gpu-vulkan-intel-windows	1d 0h 0m 1s
cpu-arm64-graviton4-kleidiai	1d 0h 0m 1s
cpu-x64-high-perf	1d 0h 0m 1s
	13d 0h 0m 10s