ContainersNVIDIADeepSeek-R1NVIDIALlama-3.1-Nemotron-70B-InstructNVIDIAPyTorch
CollectionsNVIDIAOmniverse Kit (FB)NVIDIADeepStream SDKNVIDIAOmniverse Kit App Streaming
ModelsNVIDIAStyleGAN3 pretrained modelsNVIDIAPeopleNetNVIDIATrafficCamNet
ResourcesNVIDIARiva Skills Quick Start
Helm ChartsNVIDIAGPU OperatorNVIDIANVIDIA NIM Operator
NVIDIA
NVIDIA
NeMo Framework Megatron Backend
Container
NVIDIA
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
LayerLabelCreated
sha256:1888555533c3940d2e06f20b58785c5a5a260e12a91ab103cfb20f585cb3e9fcRUN
NVIDIA_BIGNLP_VERSION= UV_VERSION=0.8.22 TRT_VER=10.13.3.9 CUDA_VER=13.0 CUDNN_VER=9.13.1.26-1 NCCL_VER=2.28.3-1+cuda13.0 CUBLAS_VER=13.0.2.14-1 TRT_LLM_COMMIT=55a9771ff0122ded47e6711eaf19c48f54630fe6 TRT_LLM_VERSION= REINSTALL_NSYS=False NSYS_VERSION=2025.4.1.172 REINSTALL_CUDNN=False CUDNN_VERSION=9.13.0.50 REINSTALL_NCCL=False NCCL_VERSION=2.27.7 INSTALL_DEEPEP=True DEEPEP_COMMIT=ef73fd9a4c098e09929e48dfd696505ddc8e2043 TARGET_ARCH=arm MCORE_REPO=https://github.com/NVIDIA/Megatron-LM.git MCORE_COMMIT=696a38e0b2e87e7bcd7412e1d45e8943f646a8d3 MCORE_VERSION=0.16.0rc0 NEMO_COMMIT=0425500334506b9b9f8dfa554fd7d920746a08cd NEMO_VERSION=2.6.0rc0 NEMO_EXPORT_DEPLOY_COMMIT=0947b64f45136550d0c570b233d370870edd4d42 NEMO_EXPORT_DEPLOY_VERSION= NEMO_EVAL_COMMIT= NEMO_EVAL_VERSION= MEGATRON_BRIDGE_COMMIT=57c3c149fbde7bcc6c36d763781bb1fa560ba0db MEGATRON_BRIDGE_VERSION=0.2.0 INSTALL_RESIL=True RESIL_COMMIT=54f85fe422d296cf04ea524130014bd3a2c3add1 RESIL_VERSION=0.4.1 INSTALL_MODELOPT=True MODELOPT_VERSION=0.37.0 NEMO_RUN_COMMIT=01a9a8ba360f7b2908728ad0516e0ad9d936966d SKIP_TRTLLM=false URM_USER_ACCOUNT=donghyukc PRE_COMPILE_QUANT_KERNELS=True NVIDIA_BUILD_ID=243847215 NVIDIA_BUILD_REF=58514fea65f0161e1a2180f788544e25b9ae8111 RC_DATE=25.11.01 TARGETARCH=arm64 echo "This distribution includes open source which is archived at the following URL: https://opensource.nvidia.com/oss/teams/nvidia/nemo/${RC_DATE}:linux-${TARGETARCH}/index.html" > NOTICES.txt &&
  echo "For further inquiries or assistance, contact us at oss-requests@nvidia.com" >> NOTICES.txt
12/12/2025 12:58 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ARG
TARGETARCH=arm64
12/12/2025 12:58 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ARG
RC_DATE=25.11.01
12/12/2025 12:58 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4LABEL
com.nvidia.build.ref=58514fea65f0161e1a2180f788544e25b9ae8111
12/12/2025 12:58 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ARG
NVIDIA_BUILD_REF=58514fea65f0161e1a2180f788544e25b9ae8111
12/12/2025 12:58 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4LABEL
com.nvidia.build.id=243847215
12/12/2025 12:58 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ENV
NVIDIA_BUILD_ID=243847215
12/12/2025 12:58 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ARG
NVIDIA_BUILD_ID=243847215
12/12/2025 12:58 AM UTC
sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1WORKDIR
/workspace
12/12/2025 12:58 AM UTC
sha256:84ca71ed3cd348d04fd1af15a14a162129d9eb9150f13b9903a6f33424ac7aafRUN
NVIDIA_BIGNLP_VERSION= UV_VERSION=0.8.22 TRT_VER=10.13.3.9 CUDA_VER=13.0 CUDNN_VER=9.13.1.26-1 NCCL_VER=2.28.3-1+cuda13.0 CUBLAS_VER=13.0.2.14-1 TRT_LLM_COMMIT=55a9771ff0122ded47e6711eaf19c48f54630fe6 TRT_LLM_VERSION= REINSTALL_NSYS=False NSYS_VERSION=2025.4.1.172 REINSTALL_CUDNN=False CUDNN_VERSION=9.13.0.50 REINSTALL_NCCL=False NCCL_VERSION=2.27.7 INSTALL_DEEPEP=True DEEPEP_COMMIT=ef73fd9a4c098e09929e48dfd696505ddc8e2043 TARGET_ARCH=arm MCORE_REPO=https://github.com/NVIDIA/Megatron-LM.git MCORE_COMMIT=696a38e0b2e87e7bcd7412e1d45e8943f646a8d3 MCORE_VERSION=0.16.0rc0 NEMO_COMMIT=0425500334506b9b9f8dfa554fd7d920746a08cd NEMO_VERSION=2.6.0rc0 NEMO_EXPORT_DEPLOY_COMMIT=0947b64f45136550d0c570b233d370870edd4d42 NEMO_EXPORT_DEPLOY_VERSION= NEMO_EVAL_COMMIT= NEMO_EVAL_VERSION= MEGATRON_BRIDGE_COMMIT=57c3c149fbde7bcc6c36d763781bb1fa560ba0db MEGATRON_BRIDGE_VERSION=0.2.0 INSTALL_RESIL=True RESIL_COMMIT=54f85fe422d296cf04ea524130014bd3a2c3add1 RESIL_VERSION=0.4.1 INSTALL_MODELOPT=True MODELOPT_VERSION=0.37.0 NEMO_RUN_COMMIT=01a9a8ba360f7b2908728ad0516e0ad9d936966d SKIP_TRTLLM=false URM_USER_ACCOUNT=donghyukc PRE_COMPILE_QUANT_KERNELS=True if [ $PRE_COMPILE_QUANT_KERNELS = "True" ]; then python -c "import modelopt.torch.quantization.extensions as ext; ext.precompile()"; fi &&
  python -m ipykernel install --name=python3 --display-name="Python 3 (ipykernel)" &&
  if [ -f /usr/local/bin/torchrun ]; then sed -i '1s|#!/usr/bin/python|#!/usr/bin/env python|' /usr/local/bin/torchrun; fi &&
  mkdir -p /workspace &&
  chmod -R a+w /workspace &&
  URM_API_TOKEN=$(cat /run/secrets/URM_API_TOKEN) &&
  wget --http-user $URM_USER_ACCOUNT --http-password ${URM_API_TOKEN} "https://urm.nvidia.com/artifactory/nemo-fw-generic-local/llama3_70b_tokenizer.tar.gz" &&
  wget --http-user $URM_USER_ACCOUNT --http-password ${URM_API_TOKEN} "https://urm.nvidia.com/artifactory/nemo-fw-generic-local/llama3_8b_tokenizer.tar.gz" &&
  mkdir -p /tmp_assets &&
  mkdir -p ~/.cache/huggingface &&
  tar -xzvf llama3_8b_tokenizer.tar.gz -C /tmp_assets &&
  tar -xzvf llama3_70b_tokenizer.tar.gz -C /tmp_assets &&
  cp -rf /tmp_assets/hub ~/.cache/huggingface &&
  rm -rf /tmp_assets &&
  rm llama3_8b_tokenizer.tar.gz llama3_70b_tokenizer.tar.gz &&
  python -c "from transformers import AutoTokenizer, Qwen2Tokenizer, CLIPImageProcessor; _=AutoTokenizer.from_pretrained('gpt2'); _=AutoTokenizer.from_pretrained('bert-base-cased'); _=AutoTokenizer.from_pretrained('bert-large-cased'); _=AutoTokenizer.from_pretrained('bert-large-uncased'); _=AutoTokenizer.from_pretrained('bigcode/starcoder2-tokenizer'); _=AutoTokenizer.from_pretrained('EleutherAI/gpt-neox-20b'); _=AutoTokenizer.from_pretrained('nvidia/Nemotron-H-8B-Base-8K'); _=AutoTokenizer.from_pretrained('nvidia/Nemotron-H-47B-Base-8K'); _=AutoTokenizer.from_pretrained('nvidia/Nemotron-H-56B-Base-8K'); _=AutoTokenizer.from_pretrained('THUDM/chatglm2-6b',trust_remote_code=True); _=AutoTokenizer.from_pretrained('THUDM/chatglm3-6b',trust_remote_code=True); _=Qwen2Tokenizer.from_pretrained('qwen/Qwen1.5-7B',trust_remote_code=True); _=Qwen2Tokenizer.from_pretrained('qwen/Qwen1.5-14B',trust_remote_code=True); _=AutoTokenizer.from_pretrained('openai/clip-vit-large-patch14'); _=AutoTokenizer.from_pretrained('meta-llama/Meta-Llama-3-8B', local_files_only=True); _=AutoTokenizer.from_pretrained('meta-llama/Meta-Llama-3-70B', local_files_only=True); _=AutoTokenizer.from_pretrained('meta-llama/Meta-Llama-3-8B-instruct', local_files_only=True); _=AutoTokenizer.from_pretrained('meta-llama/Meta-Llama-3-70B-instruct', local_files_only=True); _=CLIPImageProcessor.from_pretrained('openai/clip-vit-large-patch14'); _=CLIPImageProcessor.from_pretrained('openai/clip-vit-large-patch14-336')"
12/12/2025 12:58 AM UTC
...

NVIDIA uses cookies to improve your experience on our web site. We and our third-party partners also use cookies and other tools to collect and record information you provide as well as information about your interactions with our websites for performance improvement, analytics, and to assist in marketing efforts. By clicking "Accept All", you consent to our use of cookies and other tools as described in our Cookie Policy. You can manage your cookie settings by clicking on "Manage Settings." By continuing to use this site or by clicking one of the buttons below, you agree to our Terms of Service (which contains important waivers). Please see our Privacy Policy for more information on our privacy practices.