fix: INVALID_WORKLOAD 只跳过当前shape而不中止所有配置
原来 ADAPTIVE_FATAL_WORKLOAD=1 会设 abort_all=1,导致所有剩余TP/DP配置 全部被跳过。改为 continue 只跳过当前 shape,让后续 shape 和配置正常 继续执行。 这修复了 TP=8/DP=2 在 isl=4096 osl=128 因 KV cache 不足导致 workload 验证失败后,TP=16/DP=1 被跳过的 bug。
This commit is contained in:
parent
e53b2c7e4c
commit
ad4fd2b878
@ -945,9 +945,9 @@ PY
|
||||
adaptive_run_shape "$tp" "$dp" "$mark" "$isl" "$osl" "$config_root"
|
||||
|
||||
if (( ADAPTIVE_FATAL_WORKLOAD == 1 )); then
|
||||
log "ERROR: invalid synthetic workload; aborting remaining searches"
|
||||
abort_all=1
|
||||
break
|
||||
log "WARNING: invalid workload for tp=${tp} dp=${dp} isl=${isl} osl=${osl}; skipping shape, continuing next"
|
||||
ADAPTIVE_FATAL_WORKLOAD=0
|
||||
continue
|
||||
fi
|
||||
if (( ADAPTIVE_SKIP_CONFIG == 1 )); then
|
||||
log "config skip requested; reason=${ADAPTIVE_SKIP_CONFIG_REASON:-unknown}; skipping remaining shapes for tp=${tp} dp=${dp}"
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user