We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
1 parent 495025a commit d685275Copy full SHA for d685275
qa/L1_pytorch_mcore_fsdp_integration/test.sh
@@ -4,10 +4,10 @@
4
5
set -e
6
7
-# Megatron-LM / Megatron-FSDP commit for main branch on Apr. 7, 2026.
+# Megatron-LM / Megatron-FSDP commit for main branch on Apr. 10, 2026.
8
# Necessary to support wgrad accumulate fusion and Megatron-FSDP NCCL UBR,
9
-# and TODO(@cspades) fixes decoupled_grad <> DistOpt usage in Megatron-LM.
10
-MCORE_REF=${1:-8cbc45b6e039f300c53eb09579fc973d703455cd}
+# and fixes decoupled_grad <> DistOpt usage in Megatron-LM.
+MCORE_REF=${1:-ab43d43f0bc04f4656d4af15afb6e7e4c9ad71c8}
11
12
# Paths
13
: ${TE_PATH:=/opt/transformerengine}
0 commit comments