From acb9a0fec8d8a4dabeedcbb8e08c26f66d7083f0 Mon Sep 17 00:00:00 2001
From: haoneng.lhn <haoneng.lhn@alibaba-inc.com>
Date: 星期五, 08 十二月 2023 16:19:00 +0800
Subject: [PATCH] fix loss normalization for ddp training

---
 docs/modelscope_pipeline/quick_start.md |    5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/docs/modelscope_pipeline/quick_start.md b/docs/modelscope_pipeline/quick_start.md
index 7e35e91..2b9219b 100644
--- a/docs/modelscope_pipeline/quick_start.md
+++ b/docs/modelscope_pipeline/quick_start.md
@@ -1,3 +1,5 @@
+([绠�浣撲腑鏂嘳(./quick_start_zh.md)|English)
+
 # Quick Start
 
 > **Note**: 
@@ -221,5 +223,4 @@
 If you want finetune with multi-GPUs, you could:
 ```shell
 CUDA_VISIBLE_DEVICES=1,2 python -m torch.distributed.launch --nproc_per_node 2 finetune.py > log.txt 2>&1
-```
-
+```
\ No newline at end of file

--
Gitblit v1.9.1