bump mcore

Signed-off-by: Oliver Koenig <[email protected]>

Dockerfile.ci

@@ -49,7 +49,6 @@ RUN pip install nemo_run@git+https://github.com/NVIDIA/NeMo-Run.git@${NEMO_RUN_T
 # Install NeMo requirements
 ARG TE_TAG=7d576ed25266a17a7b651f2c12e8498f67e0baea
 ARG MODELOPT_VERSION=0.19.0
-ARG MCORE_TAG=c1728c12f1f1cdbb786e52f1ffe512295d76bef3
 
 ARG APEX_TAG=810ffae374a2b9cb4b5c5e28eaeca7d7998fca0c
 RUN \
@@ -60,25 +59,22 @@ RUN \
   --mount=type=bind,source=nemo/__init__.py,target=nemo/__init__.py <<"EOF" bash -ex
 pip install --no-cache-dir --no-build-isolation --extra-index-url https://pypi.nvidia.com \
 "transformer-engine @ git+https://github.com/NVIDIA/TransformerEngine.git@${TE_TAG}" \
-"megatron_core @ git+https://github.com/NVIDIA/Megatron-LM.git@${MCORE_TAG}" \
 "nvidia-modelopt[torch]~=${MODELOPT_VERSION}" \
 "apex @ git+https://github.com/NVIDIA/apex.git@${APEX_TAG}" \
 "unstructured==0.14.9" \
 "llama-index==0.10.43" \
 "onnxscript @ git+https://github.com/microsoft/onnxscript" \
 -r tools/ctc_segmentation/requirements.txt \
 ".[all]"
+EOF
 
-# Megatron Core installation
-git clone https://github.com/NVIDIA/Megatron-LM.git && \
-pushd Megatron-LM && \
-git checkout ${MCORE_TAG} && \
-  pushd megatron/core/datasets && \
-  make && \
-  popd && \
-popd
+ARG MCORE_TAG=452d52083081398e09fb5cdd753434aa6f1b2d1f
+RUN <<"EOF" bash -ex
+# Megatron-LM installation
+git clone https://github.com/NVIDIA/Megatron-LM.git
+pushd Megatron-LM
+git checkout ${MCORE_TAG} 
 export PYTHONPATH="${PYTHONPATH}:/workspace/Megatron-LM"
-
 EOF
 
 # Copy over NeMo code

requirements/requirements_nlp.txt

@@ -10,7 +10,7 @@ ijson
 jieba
 markdown2
 matplotlib>=3.3.2
-#megatron_core>0.6.0 # add back once mcore on pypi is compatible again
+megatron_core @ git+https://github.com/NVIDIA/Megatron-LM.git@452d52083081398e09fb5cdd753434aa6f1b2d1f
 nltk>=3.6.5
 numpy<2  # tensorstore has an implicit compiled dependency on numpy<2
 opencc