Skip to content
CI / cuda

Use fused softmax kernel in llama attention layer #3356

Sign in to view logs

Use fused softmax kernel in llama attention layer

Use fused softmax kernel in llama attention layer #3356

Re-run triggered October 23, 2024 17:43
@LaurentMazareLaurentMazare
#2572
zackangelo:llama_softmax_last_dim
Status Skipped
Total duration 4s
Artifacts –

ci_cuda.yaml

on: pull_request
test-cuda
0s
test-cuda
Fit to window
Zoom out
Zoom in