perf: frank-default-large batch 12288 / ubatch 4096

This commit is contained in:
Frank
2026-08-11 16:36:39 -07:00
parent f5a959bfa2
commit bf64c0d9c2
2 changed files with 4 additions and 4 deletions
+2 -2
View File
@@ -23,7 +23,7 @@ N_GPU_LAYERS=all
# KV on GPU, quantized (K q8_0, V f16).
CACHE_TYPE_K=q8_0
CACHE_TYPE_V=f16
BATCH_SIZE=2048
UBATCH_SIZE=512
BATCH_SIZE=12288
UBATCH_SIZE=4096
THREADS=8
THREADS_HTTP=8
+2 -2
View File
@@ -35,8 +35,8 @@ fi
: "${N_GPU_LAYERS:=all}"
: "${CACHE_TYPE_K:=q8_0}"
: "${CACHE_TYPE_V:=f16}"
: "${BATCH_SIZE:=2048}"
: "${UBATCH_SIZE:=512}"
: "${BATCH_SIZE:=12288}"
: "${UBATCH_SIZE:=4096}"
: "${THREADS:=8}"
: "${THREADS_HTTP:=8}"