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
c6d1e530b031a763620a1c1e567ae6ee2ce52bb7e4688698486780da751834b4	CONFIG	`Entrypoint /opt/nvidia/nvidia_entrypoint.sh; WorkingDir /workspace; ExposedPorts 6006/tcp, 8888/tcp`	08/28/2025 12:19 AM UTC
fc75902eef7655babe8821ddda2862e4518444b9ad7e35c00a729695b724960d	ARG	`NVIDIA_BUILD_REF`	08/28/2025 12:19 AM UTC
93549626116cccabe63eb5649abef5162ab678c57c575b77beeb45963d1ecbb6	LABEL	`com.nvidia.build.id=148941828`	08/28/2025 12:19 AM UTC
3fdfefffced3e3c0849f5e6e5de1c6661b13978e387adee7207cda83fceca705	ENV	`NVIDIA_BUILD_ID=148941828`	08/28/2025 12:19 AM UTC
6b5c93db2ae0067de250f8a8141b0471cfb9528f4fcc5738c99a4576ef567f23	ARG	`NVIDIA_BUILD_ID`	08/28/2025 12:19 AM UTC
24d7e693f7aa1875d397bf938f392236b51e8cd9ebe06a7f030439fa09d406a9	ENV	`NEMO_FW_PIPELINE_ID=33966346`	08/28/2025 12:19 AM UTC
7d73bc4bf6e5a0cfe2889da4a26ec0caf1bd1f58c3d4d29348bd20184efe6195	ARG	`NEMO_FW_PIPELINE_ID=33966346`	08/28/2025 12:19 AM UTC
8a4bb341d519775d045324b05eddec5d5cb7a5d686a5956d989830ca17376af5	RUN	NVIDIA_BIGNLP_VERSION= TRT_VER=10.9.0.34 CUDA_VER=12.8 CUDNN_VER=9.8.0.87-1 NCCL_VER=2.25.1-1+cuda12.8 CUBLAS_VER=12.8.4.1-1 NVRTC_VER=12.8.93-1 TRT_LLM_COMMIT=3c5fe5eb86077f67febc42070be11f11de17c1e2 TRT_LLM_VERSION=0.18.0 REINSTALL_NSYS=False NSYS_VERSION=2025.1.1.65 TARGET_ARCH=x86 DEEPEP_COMMIT=a84a24808fb0ea732f49b874cc456a69dde69076 GDR_COPY_VERSION=v2.4.1 REINSTALL_APEX=False APEX_COMMIT=26bba57d62553d268319b4a20cc3d8aa990249ec MCORE_COMMIT=3ea68ad6042cc1204386ae9364358f7c4de1bc37 MCORE_VERSION=0.12.3 SRC_DIR=/src/tensorrt_llm CPP_BUILD_DIR=/src/tensorrt_llm/cpp/build INSTALL_VLLM=True ALIGNER_COMMIT=dev ALIGNER_VERSION=0.6.0rc0.dev0 LAUNCHER_COMMIT=26e42fa91d9e608897ce485b4911c263a4b57008 CURATOR_COMMIT=043c67f34c70759899b2bb1f55dd797d4f75c271 INSTALL_CURATOR=True CURATOR_VERSION=0.8.1 NEMO_COMMIT=ca0af5eaa63dd4acf5f66e9a6e98d980e956efe9 NEMO_VERSION=2.3.2 INSTALL_RESIL=False RESIL_COMMIT=not_installed RESIL_VERSION=not_installed INSTALL_MODELOPT=True MODELOPT_VERSION=0.27.1 NEMO_RUN_COMMIT=bc412ee5584ed3072717af59f54565ec0d265a6f PRE_COMPILE_QUANT_KERNELS=True chmod -R a+w /workspace	08/28/2025 12:19 AM UTC
53cb6076d57b6249968204bd32ef1e5c88069b6156cca90938061e1cddb81252	WORKDIR	`/workspace`	08/28/2025 12:19 AM UTC
fe6ebf60edaa3451ef9bfd25afdb685691593f1cfcb09b30a252f10d44141242	RUN	NVIDIA_BIGNLP_VERSION= TRT_VER=10.9.0.34 CUDA_VER=12.8 CUDNN_VER=9.8.0.87-1 NCCL_VER=2.25.1-1+cuda12.8 CUBLAS_VER=12.8.4.1-1 NVRTC_VER=12.8.93-1 TRT_LLM_COMMIT=3c5fe5eb86077f67febc42070be11f11de17c1e2 TRT_LLM_VERSION=0.18.0 REINSTALL_NSYS=False NSYS_VERSION=2025.1.1.65 TARGET_ARCH=x86 DEEPEP_COMMIT=a84a24808fb0ea732f49b874cc456a69dde69076 GDR_COPY_VERSION=v2.4.1 REINSTALL_APEX=False APEX_COMMIT=26bba57d62553d268319b4a20cc3d8aa990249ec MCORE_COMMIT=3ea68ad6042cc1204386ae9364358f7c4de1bc37 MCORE_VERSION=0.12.3 SRC_DIR=/src/tensorrt_llm CPP_BUILD_DIR=/src/tensorrt_llm/cpp/build INSTALL_VLLM=True ALIGNER_COMMIT=dev ALIGNER_VERSION=0.6.0rc0.dev0 LAUNCHER_COMMIT=26e42fa91d9e608897ce485b4911c263a4b57008 CURATOR_COMMIT=043c67f34c70759899b2bb1f55dd797d4f75c271 INSTALL_CURATOR=True CURATOR_VERSION=0.8.1 NEMO_COMMIT=ca0af5eaa63dd4acf5f66e9a6e98d980e956efe9 NEMO_VERSION=2.3.2 INSTALL_RESIL=False RESIL_COMMIT=not_installed RESIL_VERSION=not_installed INSTALL_MODELOPT=True MODELOPT_VERSION=0.27.1 NEMO_RUN_COMMIT=bc412ee5584ed3072717af59f54565ec0d265a6f PRE_COMPILE_QUANT_KERNELS=True cd /usr/local/lib/python3.12/dist-packages && patch -p1 < /opt/triton-lang_triton_6570_lazy_init.patch && rm /opt/triton-lang_triton_6570_lazy_init.patch	08/28/2025 12:19 AM UTC