From 990b5896bc8abe11b6cce0eee0edf2629f85d9b6 Mon Sep 17 00:00:00 2001 From: NanoCode012 Date: Tue, 1 Apr 2025 23:25:05 +0700 Subject: [PATCH] fix: downgrade deepspeed to fix grad checkpoint oom (#2465) [skip ci] --- requirements.txt | 2 +- setup.py | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/requirements.txt b/requirements.txt index 567b446dd..dde64f392 100644 --- a/requirements.txt +++ b/requirements.txt @@ -16,7 +16,7 @@ transformers==4.50.3 tokenizers>=0.21.1 accelerate==1.5.2 datasets==3.5.0 -deepspeed==0.16.4 +deepspeed==0.15.4 trl==0.16.0 optimum==1.16.2 diff --git a/setup.py b/setup.py index aef8182af..29719b1f3 100644 --- a/setup.py +++ b/setup.py @@ -112,7 +112,7 @@ extras_require = { "yunchang==0.6.0", ], "deepspeed": [ - "deepspeed==0.16.4", + "deepspeed==0.15.4", "deepspeed-kernels", ], "mamba-ssm": [