Skip to main content
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
nemo-rl
Container
NVIDIA
nemo-rl

NVIDIA NeMo™ RL accelerates reinforcement learning post-training with high-performance GPU backends, offering scalable GRPO, DPO, SFT, and distillation for multimodal models from single-node experiments to enterprise-scale clusters.

LayerLabelCreated
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4USER
root
11/05/2026 9:58 PM UTC
sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1RUN
RUN |37 UV_VERSION=0.11.28 PYTHON_VERSION=3.13.14 MAX_JOBS= NVTE_BUILD_MAX_JOBS=8 NVTE_BUILD_THREADS_PER_JOB=2 NVTE_CUDA_ARCHS=90;100 NVTE_FRAMEWORK=pytorch NVTE_WITH_NCCL_EP=0 BUILD_CUSTOM_VLLM= BUILD_CUSTOM_VLLM_URL= BUILD_CUSTOM_VLLM_REF= BUILD_CUSTOM_VLLM_PRECOMPILED_WHEEL_LOCATION= BUILD_CUSTOM_FLASHINFER= BUILD_CUSTOM_FLASHINFER_URL= BUILD_CUSTOM_FLASHINFER_REF= SKIP_VLLM_BUILD= SKIP_SGLANG_BUILD= SKIP_TRTLLM_BUILD= BASE_IMAGE=nvcr.io/nvidia/cuda-dl-base:26.05-cuda13.2-devel-ubuntu24.04 TARGETARCH=arm64 TRTLLM_CCACHE_MAXSIZE=20G STRIP_UV_CACHE= TRTLLM_BUILD_JOBS=24 CUSTOM_SETUP_FNAME=install_apptainer.sh BUILD_DYNAMO= DYNAMO_PYTHON_VERSION=3.12.11 ETCD_VERSION=v3.5.21 NATS_VERSION=v2.11.6 NEMO_GYM_PREFETCH_CONFIGS= NEMO_RL_COMMIT=a09162e1e66b292af6e0b188a685af69a6fe408e NVIDIA_BUILD_ID=530887851 NVIDIA_BUILD_REF=3ea32d14ec1477ba463714d2467563468ce65002 RC_DATE=v0.7.0 NEMO_GYM_CUDA=cu130 NEMO_GYM_VLLM_VERSION=0.25.1 RUNTIME_UID=65532 RUNTIME_GID=65532 /bin/sh -c <<"EOF" bash -exu uv --version "${UV_PROJECT_ENVIRONMENT}/bin/python" --version for dir in /home/nemo-runtime /nemo_run /opt/nemo-rl "${UV_PROJECT_ENVIRONMENT}" "${NEMO_RL_VENV_DIR}" \
"${NEMO_GYM_VENV_DIR}" /opt/nemo-rl-source /opt/nemo_rl_cache/uv /opt/nemo_rl_python; do test -w "${dir}" done EOF # buildkit
11/05/2026 9:58 PM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4USER
65532:65532
11/05/2026 9:58 PM UTC
sha256:2d57eafe8fb4c5f31a5d4ee9b1645bd50d7b86c516e64bf59500167eb3bb71eaRUN
RUN |37 UV_VERSION=0.11.28 PYTHON_VERSION=3.13.14 MAX_JOBS= NVTE_BUILD_MAX_JOBS=8 NVTE_BUILD_THREADS_PER_JOB=2 NVTE_CUDA_ARCHS=90;100 NVTE_FRAMEWORK=pytorch NVTE_WITH_NCCL_EP=0 BUILD_CUSTOM_VLLM= BUILD_CUSTOM_VLLM_URL= BUILD_CUSTOM_VLLM_REF= BUILD_CUSTOM_VLLM_PRECOMPILED_WHEEL_LOCATION= BUILD_CUSTOM_FLASHINFER= BUILD_CUSTOM_FLASHINFER_URL= BUILD_CUSTOM_FLASHINFER_REF= SKIP_VLLM_BUILD= SKIP_SGLANG_BUILD= SKIP_TRTLLM_BUILD= BASE_IMAGE=nvcr.io/nvidia/cuda-dl-base:26.05-cuda13.2-devel-ubuntu24.04 TARGETARCH=arm64 TRTLLM_CCACHE_MAXSIZE=20G STRIP_UV_CACHE= TRTLLM_BUILD_JOBS=24 CUSTOM_SETUP_FNAME=install_apptainer.sh BUILD_DYNAMO= DYNAMO_PYTHON_VERSION=3.12.11 ETCD_VERSION=v3.5.21 NATS_VERSION=v2.11.6 NEMO_GYM_PREFETCH_CONFIGS= NEMO_RL_COMMIT=a09162e1e66b292af6e0b188a685af69a6fe408e NVIDIA_BUILD_ID=530887851 NVIDIA_BUILD_REF=3ea32d14ec1477ba463714d2467563468ce65002 RC_DATE=v0.7.0 NEMO_GYM_CUDA=cu130 NEMO_GYM_VLLM_VERSION=0.25.1 RUNTIME_UID=65532 RUNTIME_GID=65532 /bin/sh -c <<"EOF" bash -exu getent group "${RUNTIME_GID}" >/dev/null || groupadd --gid "${RUNTIME_GID}" nemo-runtime getent passwd "${RUNTIME_UID}" >/dev/null || useradd --no-log-init --uid "${RUNTIME_UID}" \
--gid "${RUNTIME_GID}" --create-home --home-dir /home/nemo-runtime --shell /bin/bash nemo-runtime install -d -o "${RUNTIME_UID}" -g "${RUNTIME_GID}" /home/nemo-runtime /nemo_run \
"${NEMO_GYM_VENV_DIR}" /opt/nemo-rl-source /opt/nemo_rl_cache/uv /opt/nemo_rl_python chown "${RUNTIME_UID}:${RUNTIME_GID}" /opt/nemo-rl "${UV_PROJECT_ENVIRONMENT}" "${NEMO_RL_VENV_DIR}" EOF # buildkit
11/05/2026 9:58 PM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ARG
RUNTIME_GID=65532
11/05/2026 9:58 PM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ARG
RUNTIME_UID=65532
11/05/2026 9:58 PM UTC
sha256:16c19d0caa085e0839135312062f035ee725a23285c20f23a6321c260def06deRUN
UV_VERSION=0.11.28 PYTHON_VERSION=3.13.14 MAX_JOBS= NVTE_BUILD_MAX_JOBS=8 NVTE_BUILD_THREADS_PER_JOB=2 NVTE_CUDA_ARCHS=90;100 NVTE_FRAMEWORK=pytorch NVTE_WITH_NCCL_EP=0 BUILD_CUSTOM_VLLM= BUILD_CUSTOM_VLLM_URL= BUILD_CUSTOM_VLLM_REF= BUILD_CUSTOM_VLLM_PRECOMPILED_WHEEL_LOCATION= BUILD_CUSTOM_FLASHINFER= BUILD_CUSTOM_FLASHINFER_URL= BUILD_CUSTOM_FLASHINFER_REF= SKIP_VLLM_BUILD= SKIP_SGLANG_BUILD= SKIP_TRTLLM_BUILD= BASE_IMAGE=nvcr.io/nvidia/cuda-dl-base:26.05-cuda13.2-devel-ubuntu24.04 TARGETARCH=arm64 TRTLLM_CCACHE_MAXSIZE=20G STRIP_UV_CACHE= TRTLLM_BUILD_JOBS=24 CUSTOM_SETUP_FNAME=install_apptainer.sh BUILD_DYNAMO= DYNAMO_PYTHON_VERSION=3.12.11 ETCD_VERSION=v3.5.21 NATS_VERSION=v2.11.6 NEMO_GYM_PREFETCH_CONFIGS= NEMO_RL_COMMIT=a09162e1e66b292af6e0b188a685af69a6fe408e NVIDIA_BUILD_ID=530887851 NVIDIA_BUILD_REF=3ea32d14ec1477ba463714d2467563468ce65002 RC_DATE=v0.7.0 NEMO_GYM_CUDA=cu130 NEMO_GYM_VLLM_VERSION=0.25.1 echo "This distribution includes open source which is archived at the following URL: https://opensource.nvidia.com/oss/teams/nvidia/nemo-rl/${RC_DATE}:linux-${TARGETARCH}/index.html" > NOTICES.txt &&
  echo "For further inquiries or assistance, contact us at oss-requests@nvidia.com" >> NOTICES.txt
11/05/2026 9:58 PM UTC
sha256:709040f2c64112f20fa5e506cdedbf3e480e1b59113526ca1463d58983891b3fRUN
UV_VERSION=0.11.28 PYTHON_VERSION=3.13.14 MAX_JOBS= NVTE_BUILD_MAX_JOBS=8 NVTE_BUILD_THREADS_PER_JOB=2 NVTE_CUDA_ARCHS=90;100 NVTE_FRAMEWORK=pytorch NVTE_WITH_NCCL_EP=0 BUILD_CUSTOM_VLLM= BUILD_CUSTOM_VLLM_URL= BUILD_CUSTOM_VLLM_REF= BUILD_CUSTOM_VLLM_PRECOMPILED_WHEEL_LOCATION= BUILD_CUSTOM_FLASHINFER= BUILD_CUSTOM_FLASHINFER_URL= BUILD_CUSTOM_FLASHINFER_REF= SKIP_VLLM_BUILD= SKIP_SGLANG_BUILD= SKIP_TRTLLM_BUILD= BASE_IMAGE=nvcr.io/nvidia/cuda-dl-base:26.05-cuda13.2-devel-ubuntu24.04 TARGETARCH=arm64 TRTLLM_CCACHE_MAXSIZE=20G STRIP_UV_CACHE= TRTLLM_BUILD_JOBS=24 CUSTOM_SETUP_FNAME=install_apptainer.sh BUILD_DYNAMO= DYNAMO_PYTHON_VERSION=3.12.11 ETCD_VERSION=v3.5.21 NATS_VERSION=v2.11.6 NEMO_GYM_PREFETCH_CONFIGS= NEMO_RL_COMMIT=a09162e1e66b292af6e0b188a685af69a6fe408e NVIDIA_BUILD_ID=530887851 NVIDIA_BUILD_REF=3ea32d14ec1477ba463714d2467563468ce65002 RC_DATE=v0.7.0 NEMO_GYM_CUDA=cu130 NEMO_GYM_VLLM_VERSION=0.25.1 python tools/generate_fingerprint.py > /opt/nemo_rl_container_fingerprint
11/05/2026 9:58 PM UTC
sha256:205c9fd9661af8f1e5d6ddc0e399f71449f30703e672e800947880c15b19f2a1RUN
UV_VERSION=0.11.28 PYTHON_VERSION=3.13.14 MAX_JOBS= NVTE_BUILD_MAX_JOBS=8 NVTE_BUILD_THREADS_PER_JOB=2 NVTE_CUDA_ARCHS=90;100 NVTE_FRAMEWORK=pytorch NVTE_WITH_NCCL_EP=0 BUILD_CUSTOM_VLLM= BUILD_CUSTOM_VLLM_URL= BUILD_CUSTOM_VLLM_REF= BUILD_CUSTOM_VLLM_PRECOMPILED_WHEEL_LOCATION= BUILD_CUSTOM_FLASHINFER= BUILD_CUSTOM_FLASHINFER_URL= BUILD_CUSTOM_FLASHINFER_REF= SKIP_VLLM_BUILD= SKIP_SGLANG_BUILD= SKIP_TRTLLM_BUILD= BASE_IMAGE=nvcr.io/nvidia/cuda-dl-base:26.05-cuda13.2-devel-ubuntu24.04 TARGETARCH=arm64 TRTLLM_CCACHE_MAXSIZE=20G STRIP_UV_CACHE= TRTLLM_BUILD_JOBS=24 CUSTOM_SETUP_FNAME=install_apptainer.sh BUILD_DYNAMO= DYNAMO_PYTHON_VERSION=3.12.11 ETCD_VERSION=v3.5.21 NATS_VERSION=v2.11.6 NEMO_GYM_PREFETCH_CONFIGS= NEMO_RL_COMMIT=a09162e1e66b292af6e0b188a685af69a6fe408e NVIDIA_BUILD_ID=530887851 NVIDIA_BUILD_REF=3ea32d14ec1477ba463714d2467563468ce65002 RC_DATE=v0.7.0 NEMO_GYM_CUDA=cu130 NEMO_GYM_VLLM_VERSION=0.25.1 find /opt /root -name ray_dist.jar -delete
11/05/2026 9:58 PM UTC
sha256:138eef30233bea8263173bafff10f12ffb2aad5e6e115c7a4ce8a70f1752ec90RUN
RUN |35 UV_VERSION=0.11.28 PYTHON_VERSION=3.13.14 MAX_JOBS= NVTE_BUILD_MAX_JOBS=8 NVTE_BUILD_THREADS_PER_JOB=2 NVTE_CUDA_ARCHS=90;100 NVTE_FRAMEWORK=pytorch NVTE_WITH_NCCL_EP=0 BUILD_CUSTOM_VLLM= BUILD_CUSTOM_VLLM_URL= BUILD_CUSTOM_VLLM_REF= BUILD_CUSTOM_VLLM_PRECOMPILED_WHEEL_LOCATION= BUILD_CUSTOM_FLASHINFER= BUILD_CUSTOM_FLASHINFER_URL= BUILD_CUSTOM_FLASHINFER_REF= SKIP_VLLM_BUILD= SKIP_SGLANG_BUILD= SKIP_TRTLLM_BUILD= BASE_IMAGE=nvcr.io/nvidia/cuda-dl-base:26.05-cuda13.2-devel-ubuntu24.04 TARGETARCH=arm64 TRTLLM_CCACHE_MAXSIZE=20G STRIP_UV_CACHE= TRTLLM_BUILD_JOBS=24 CUSTOM_SETUP_FNAME=install_apptainer.sh BUILD_DYNAMO= DYNAMO_PYTHON_VERSION=3.12.11 ETCD_VERSION=v3.5.21 NATS_VERSION=v2.11.6 NEMO_GYM_PREFETCH_CONFIGS= NEMO_RL_COMMIT=a09162e1e66b292af6e0b188a685af69a6fe408e NVIDIA_BUILD_ID=530887851 NVIDIA_BUILD_REF=3ea32d14ec1477ba463714d2467563468ce65002 RC_DATE=v0.7.0 NEMO_GYM_CUDA=cu130 NEMO_GYM_VLLM_VERSION=0.25.1 /bin/sh -c <<"EOF" bash -exu if [[ -n "${NEMO_GYM_PREFETCH_CONFIGS:-}" ]]; then case "${TARGETARCH:-amd64}" in arm64) WHEEL_ARCH=aarch64 ;; *) WHEEL_ARCH=x86_64 ;; esac GYM_VLLM_OVERRIDE=/opt/nemo-rl/.gym-vllm-override.txt printf 'vllm @ https://github.com/vllm-project/vllm/releases/download/v%s/vllm-%s-cp38-abi3-manylinux_2_28_%s.whl\n' \
"${NEMO_GYM_VLLM_VERSION}" "${NEMO_GYM_VLLM_VERSION}" "${WHEEL_ARCH}" > "${GYM_VLLM_OVERRIDE}" UV_TORCH_BACKEND="${NEMO_GYM_CUDA}" \
UV_OVERRIDE="${GYM_VLLM_OVERRIDE}" \
UV_LINK_MODE=symlink uv run python examples/nemo_gym/prefetch_venvs.py $NEMO_GYM_PREFETCH_CONFIGS rm -f "${GYM_VLLM_OVERRIDE}" fi # Release-stage editable installs and Gym prefetch can cache old build-only # setuptools again. Prune after the final uv install, not just backend warmups. bash /usr/local/libexec/prune_setuptools_cache.sh /root/.cache/uv EOF # buildkit
11/05/2026 9:58 PM UTC
1-10 of 168 items

Modal Content

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.

Manage Settings