Skip to content
Server

fix: allow mixed turbo/q8_0 KV cache types in CUDA flash-attention selection #11

Sign in to view logs

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

fix: allow mixed turbo/q8_0 KV cache types in CUDA flash-attention selection #11

Triggered via pull request July 18, 2026 22:28
@thecodacusthecodacus
synchronize #3
fable5/turbo-fattn-vec-routing
Status Cancelled
Total duration 2m 10s
Artifacts –

server.yml

on: pull_request
ubuntu
1m 46s
ubuntu
windows
1m 43s
windows
Fit to window
Zoom out
Zoom in

Annotations

5 errors
windows
Canceling since a higher priority waiting request for Server-refs/pull/3/merge-fable5/turbo-fattn-vec-routing exists
windows
The operation was canceled.
Server
Canceling since a higher priority waiting request for Server-refs/pull/3/merge-fable5/turbo-fattn-vec-routing exists
ubuntu
Canceling since a higher priority waiting request for Server-refs/pull/3/merge-fable5/turbo-fattn-vec-routing exists
ubuntu
The operation was canceled.