NeMo Framework Megatron Backend

NVIDIA

Container

NVIDIA

NeMo Framework Megatron Backend

NVIDIA NeMo™ framework Megatron backend supports pre-training, post-training, and reinforcement learning of LLMs and multi-modal generative AI models with state-of-the-art data processing, model training techniques, and flexible deployment options.

NVIDIA AI Enterprise Supported

Layer	Label		Created
e55e799fc9f0fe816bcecc72a942dc3982f76ad1c7b3c880e4c971cf4c60161d	CONFIG	`Entrypoint /opt/nvidia/nvidia_entrypoint.sh; WorkingDir /workspace; ExposedPorts 6006/tcp, 8888/tcp`	02/26/2025 10:56 AM UTC
799a9cad31bed3316e178defc8e34b9ac573df2d9c0714ad6b0056f6c7f346e2	ARG	`NVIDIA_BUILD_REF`	02/26/2025 10:56 AM UTC
ce3c83151e984379c420f93b232b20ab6605514bc583d12939710125f8535916	LABEL	`com.nvidia.build.id=134983853`	02/26/2025 10:56 AM UTC
0bc02022952603e240c02577e6c06b18108154b20385083d731562764c9b4fd7	ENV	`NVIDIA_BUILD_ID=134983853`	02/26/2025 10:56 AM UTC
7d23a7938c3f8688fe4eaad334222f9bb068664f3264cb6d8b8b8b6f9b86679f	ARG	`NVIDIA_BUILD_ID`	02/26/2025 10:56 AM UTC
7fe777329d96080f7628dce2bb0abb28753eb8351974e2ca4d212a8c747e97ba	ENV	`NEMO_FW_PIPELINE_ID=24597396`	02/26/2025 10:56 AM UTC
f0e4f47920d5cda8490fb1d4dd523b6ecbac44360201014728b582a1739c57b0	ARG	`NEMO_FW_PIPELINE_ID=24597396`	02/26/2025 10:56 AM UTC
80897e77099e3bc28c0196a9a43723760a3a4016d9e39ee2f4918d27b6edac07	RUN	NVIDIA_BIGNLP_VERSION= TRT_VER=10.8.0.24 CUDA_VER=12.8 CUDNN_VER=9.5.1.17-1 NCCL_VER= CUBLAS_VER=12.6.1.4-1 NVRTC_VER=12.6.85-1 TRT_LLM_COMMIT=2830e4e8099c0d7e6402494865b9cb7433f51eb5 TRT_LLM_VERSION=0.17.0 REINSTALL_APEX=False APEX_COMMIT=b216eeee7dd91478745496919e9c0167cc4c41e3 MCORE_COMMIT=791b70e031dc6905f77e29c25f1e2ac5b8dd4870 MCORE_VERSION=0.11.0rc1 SRC_DIR=/src/tensorrt_llm CPP_BUILD_DIR=/src/tensorrt_llm/cpp/build TARGET_ARCH=x86 INSTALL_VLLM=True ALIGNER_COMMIT=r0.7.0 ALIGNER_VERSION=0.7.0rc1.dev0 LAUNCHER_COMMIT=4abd481402adc8a061942c486eda6d71f19de718 DATA_CURATOR_COMMIT=r0.7.0 INSTALL_DATA_CURATOR=True DATA_CURATOR_VERSION=0.7.0rc2.dev0 INSTALL_MODELOPT=False MODELOPT_VERSION= NEMO_COMMIT=r2.2.0 NEMO_VERSION=2.2.0rc4 INSTALL_RESIL=True RESIL_COMMIT=b6eb61dbf9fe272b1a943b1b0d9efdde99df0737 RESIL_VERSION=0.2.0 NEMO_RUN_COMMIT=451f781c67698582b7ba91ac86aac0e5b5c6ae52 PRE_COMPILE_QUANT_KERNELS=False chmod -R a+w /workspace	02/26/2025 10:56 AM UTC
caa119317fe21e4aac60080f1f5cac6f6c76ce2b17b65d973142680afecfe9ab	WORKDIR	`/workspace`	02/26/2025 10:56 AM UTC
5df0a6057db70f33725865e296dceb8fe550875062c7197880205710df00a431	RUN	NVIDIA_BIGNLP_VERSION= TRT_VER=10.8.0.24 CUDA_VER=12.8 CUDNN_VER=9.5.1.17-1 NCCL_VER= CUBLAS_VER=12.6.1.4-1 NVRTC_VER=12.6.85-1 TRT_LLM_COMMIT=2830e4e8099c0d7e6402494865b9cb7433f51eb5 TRT_LLM_VERSION=0.17.0 REINSTALL_APEX=False APEX_COMMIT=b216eeee7dd91478745496919e9c0167cc4c41e3 MCORE_COMMIT=791b70e031dc6905f77e29c25f1e2ac5b8dd4870 MCORE_VERSION=0.11.0rc1 SRC_DIR=/src/tensorrt_llm CPP_BUILD_DIR=/src/tensorrt_llm/cpp/build TARGET_ARCH=x86 INSTALL_VLLM=True ALIGNER_COMMIT=r0.7.0 ALIGNER_VERSION=0.7.0rc1.dev0 LAUNCHER_COMMIT=4abd481402adc8a061942c486eda6d71f19de718 DATA_CURATOR_COMMIT=r0.7.0 INSTALL_DATA_CURATOR=True DATA_CURATOR_VERSION=0.7.0rc2.dev0 INSTALL_MODELOPT=False MODELOPT_VERSION= NEMO_COMMIT=r2.2.0 NEMO_VERSION=2.2.0rc4 INSTALL_RESIL=True RESIL_COMMIT=b6eb61dbf9fe272b1a943b1b0d9efdde99df0737 RESIL_VERSION=0.2.0 NEMO_RUN_COMMIT=451f781c67698582b7ba91ac86aac0e5b5c6ae52 PRE_COMPILE_QUANT_KERNELS=False if [ $PRE_COMPILE_QUANT_KERNELS = "True" ]; then python -c "import modelopt.torch.quantization.extensions as ext; ext.precompile()"; fi	02/26/2025 10:56 AM UTC