Commit History
14 days ago
-
Add auto arena mode (80% of available memory, capped at all experts); patch 369ef971
Verified signature2873e19 -
Hot/cold expert analysis: skew, profile pinning hit rates; fix hit counting and default pin fraction
Verified signature55dfdd3 -
Upload v2 memory tiers (default and forced row-parallel) with path counters
Verified signature47b5113
15 days ago
-
Verify EPQ against native llama.cpp; fix swiglu clamp and dump hook; re-run memory tiers
Verified signature632dc49 -
Add Flint expert paging port and memory benchmark
Verified signaturee2e6c43