|
|
@@ -42,7 +42,7 @@ shm_size = "16g"
|
|
|
# Recovery if a boot ever wedges: HA plug cycle of switch.dgx3/dgx4 ONLY
|
|
|
# (never the shared fabric switch).
|
|
|
args = [
|
|
|
- "--served-model-name", "glm53-fp4-local",
|
|
|
+ "--served-model-name", "glm53",
|
|
|
"--dtype", "bfloat16",
|
|
|
"--mamba-cache-mode", "align",
|
|
|
"--enable-prefix-caching",
|