层1: 修复远程 sanity OOM——per_device batch 8→2、累积 2→8(全局 64 不变)
根因:150k 大词表下显存大头是 (B,T,V) logits 链(fp32 ~20G@B=8)与逐层激活, 均正比于 B 而与 0.6B 参数量无关。docs/02 §2.6 旧显存估算勘误入档; train_sft.sh 加 expandable_segments 防碎片。 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Vendored
+4
@@ -0,0 +1,4 @@
|
||||
{
|
||||
"python-envs.defaultEnvManager": "ms-python.python:conda",
|
||||
"python-envs.defaultPackageManager": "ms-python.python:conda"
|
||||
}
|
||||
Reference in New Issue
Block a user