Skip to main content
Google
Gemma 4 31B IT
Container
Google
Gemma 4 31B IT

Gemma 4 31B IT model which, is an open multimodal model built by Google DeepMind that handles text and image inputs, can process video as sequences of frames, and generates text output.

LayerLabelCreated
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4CMD
["/bin/bash" "-c" "$SERVER_START_SCRIPT_PATH"]
05/28/2026 10:29 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ENTRYPOINT
—
05/28/2026 10:29 AM UTC
sha256:658ec8f5294386340b1cedadf303fc55b86a3f852ea48ac4b1187fe0295f3a5fRUN
RUN |4 TARGETARCH=amd64 TELEMETRY_MODE=0 TELEMETRY_ENDPOINT=0 PIP_INDEX_URL=https://pypi.org/simple /bin/sh -c touch $SERVER_START_SCRIPT_PATH &&
  chmod a+rx $SERVER_START_SCRIPT_PATH &&
  cat > $SERVER_START_SCRIPT_PATH <<-EOF #!/usr/bin/env bash set -eu # SPDX-FileCopyrightText: Copyright (c) 2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved. # SPDX-License-Identifier: LicenseRef-NvidiaProprietary start_server EOF # buildkit
05/28/2026 10:29 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ENV
SERVER_START_SCRIPT_PATH=/opt/nim/start_server.sh
05/28/2026 10:29 AM UTC
sha256:f905d7f937ec87d14fc0e0ea86feb382652794e5aea39f7e76716a53c6e74a46COPY
./tool_chat_template_gemma4.jinja /opt/nim/tool_chat_template_gemma4.jinja
05/28/2026 10:29 AM UTC
sha256:9539e0703121e2531809de77a0b1a2e00bbd72e7a94027feec02caa3494f1090COPY
model_manifest_prod.yaml /opt/nim/etc/default/model_manifest.yaml
05/28/2026 10:29 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4USER
nvs:1000
05/28/2026 10:29 AM UTC
sha256:3e7869de457e42a4a20d63a720b8eef838a6ae302d42ddb9064ae62ee7d2c29dRUN
TARGETARCH=amd64 TELEMETRY_MODE=0 TELEMETRY_ENDPOINT=0 PIP_INDEX_URL=https://pypi.org/simple [ -d /etc/ld.so.conf.d ] ||
  mkdir -p /etc/ld.so.conf.d &&
  [ -f /etc/ld.so.conf.d/nvidia.conf ] ||
  touch /etc/ld.so.conf.d/nvidia.conf &&
  echo "/usr/local/nvidia/lib" >> /etc/ld.so.conf.d/nvidia.conf &&
  echo "/usr/local/nvidia/lib64" >> /etc/ld.so.conf.d/nvidia.conf &&
  echo "/usr/local/cuda/lib64" >> /etc/ld.so.conf.d/nvidia.conf &&
  if command -v ldconfig >/dev/null 2>&1; then ldconfig; fi
05/28/2026 10:29 AM UTC
sha256:bb352aef230b0202e732aa24989e8b444693c33c2bc4727fd35fad666d098231RUN
TARGETARCH=amd64 TELEMETRY_MODE=0 TELEMETRY_ENDPOINT=0 PIP_INDEX_URL=https://pypi.org/simple groupadd --gid 1000 --non-unique nvs &&
  useradd --create-home --shell /usr/sbin/nologin --uid 1000 --non-unique --gid 1000 nvs &&
  chown 1000.1000 $NIM_DIR_PATH
05/28/2026 10:29 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ENV
VLLM_VIDEO_LOADER_BACKEND=pynvvideocodec
05/28/2026 10:29 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.