NVIDIA
NVIDIA
JAX
Container
NVIDIA
NVIDIA
JAX

JAX is a framework for high-performance numerical computing and machine learning research. It includes Numpy-like APIs, automatic differentiation, XLA acceleration and simple primitives for scaling across GPUs and supports an ecosystem of libraries.

LayerLabelCreated
sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1WORKDIR
/opt/maxtext
01/22/2025 8:57 AM UTC
sha256:413cba0ce52ff3db2658772f1d35bedee51456a2ed7f69527d2de7de45a795e8RUN
NVIDIA_BUILD_ID=135630783 URLREF_MAXTEXT=https://github.com/google/maxtext.git#4651cb3c73de92e528a98281be29c05b27aaf501 URLREF_TFTEXT=https://github.com/tensorflow/text.git#master SRC_PATH_MAXTEXT=/opt/maxtext SRC_PATH_TFTEXT=/opt/tensorflow-text URLREF_JETSTREAM=https://github.com/google/jetstream.git#9ca442137fec58bab819b7e561f0d204ebda4767 SRC_PATH_JETSTREAM=/opt/jetstream pip install "ml_dtypes>=0.5.0"
01/22/2025 8:57 AM UTC
sha256:2ed6413daaf5a946eb4382f5b2cffc8b5fd5daa0814f9f5c8b3db3eae30a93caRUN
NVIDIA_BUILD_ID=135630783 URLREF_MAXTEXT=https://github.com/google/maxtext.git#4651cb3c73de92e528a98281be29c05b27aaf501 URLREF_TFTEXT=https://github.com/tensorflow/text.git#master SRC_PATH_MAXTEXT=/opt/maxtext SRC_PATH_TFTEXT=/opt/tensorflow-text URLREF_JETSTREAM=https://github.com/google/jetstream.git#9ca442137fec58bab819b7e561f0d204ebda4767 SRC_PATH_JETSTREAM=/opt/jetstream pip install "werkzeug>=3.0.3"
01/22/2025 8:57 AM UTC
sha256:400a5b02a06b9df2070e381163d1779765580fc3c20ce2fdccedf138bcb1e246RUN
NVIDIA_BUILD_ID=135630783 URLREF_MAXTEXT=https://github.com/google/maxtext.git#4651cb3c73de92e528a98281be29c05b27aaf501 URLREF_TFTEXT=https://github.com/tensorflow/text.git#master SRC_PATH_MAXTEXT=/opt/maxtext SRC_PATH_TFTEXT=/opt/tensorflow-text URLREF_JETSTREAM=https://github.com/google/jetstream.git#9ca442137fec58bab819b7e561f0d204ebda4767 SRC_PATH_JETSTREAM=/opt/jetstream pip-finalize.sh
01/22/2025 8:57 AM UTC
sha256:908a787bbd53fa8964ed27c6274ffaef8f7434f7b663414491232852beec7aa2RUN
NVIDIA_BUILD_ID=135630783 URLREF_MAXTEXT=https://github.com/google/maxtext.git#4651cb3c73de92e528a98281be29c05b27aaf501 URLREF_TFTEXT=https://github.com/tensorflow/text.git#master SRC_PATH_MAXTEXT=/opt/maxtext SRC_PATH_TFTEXT=/opt/tensorflow-text URLREF_JETSTREAM=https://github.com/google/jetstream.git#9ca442137fec58bab819b7e561f0d204ebda4767 SRC_PATH_JETSTREAM=/opt/jetstream sed -i "s/ml_dtypes>=0.5.0/ml_dtypes>=0.4.0/" /opt/jax/build/requirements.in &&
  sed -i "s/ml_dtypes>=0.5.0/ml_dtypes>=0.4.0/" /opt/jax/setup.py
01/22/2025 8:53 AM UTC
sha256:8d0bd43983585f2ee0ffe99c9c1f5da852ab4e04b920821da9c9ebd5a2c196dfRUN
RUN |7 NVIDIA_BUILD_ID=135630783 URLREF_MAXTEXT=https://github.com/google/maxtext.git#4651cb3c73de92e528a98281be29c05b27aaf501 URLREF_TFTEXT=https://github.com/tensorflow/text.git#master SRC_PATH_MAXTEXT=/opt/maxtext SRC_PATH_TFTEXT=/opt/tensorflow-text URLREF_JETSTREAM=https://github.com/google/jetstream.git#9ca442137fec58bab819b7e561f0d204ebda4767 SRC_PATH_JETSTREAM=/opt/jetstream /bin/sh -c <<"EOF" bash -ex git-clone.sh ${URLREF_JETSTREAM} ${SRC_PATH_JETSTREAM} echo "-e file://${SRC_PATH_JETSTREAM}" >> /opt/pip-tools.d/requirements-jetstream.in EOF # buildkit
01/22/2025 8:53 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ARG
SRC_PATH_JETSTREAM=/opt/jetstream
01/22/2025 8:53 AM UTC
sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4ARG
URLREF_JETSTREAM=https://github.com/google/jetstream.git#9ca442137fec58bab819b7e561f0d204ebda4767
01/22/2025 8:53 AM UTC
sha256:2cfe3a370f5801426ac008a2caaa6f1ce92a11ff0cd837341477ab4224d0104dADD
test-maxtext.sh /usr/local/bin
01/22/2025 8:53 AM UTC
sha256:358fbfb6167b935dea3b47b68f0520c1b7c248dc620b87cb9df7b7af4f754810RUN
RUN |5 NVIDIA_BUILD_ID=135630783 URLREF_MAXTEXT=https://github.com/google/maxtext.git#4651cb3c73de92e528a98281be29c05b27aaf501 URLREF_TFTEXT=https://github.com/tensorflow/text.git#master SRC_PATH_MAXTEXT=/opt/maxtext SRC_PATH_TFTEXT=/opt/tensorflow-text /bin/sh -c <<"EOF" bash -ex git-clone.sh ${URLREF_MAXTEXT} ${SRC_PATH_MAXTEXT} sed -i '/google-jetstream/d' ${SRC_PATH_MAXTEXT}/requirements.txt echo "-r ${SRC_PATH_MAXTEXT}/requirements.txt" >> /opt/pip-tools.d/requirements-maxtext.in # specify some restrictions to speed up the build and # avoid pip to download and check all available versions of packages for pattern in \
"s|absl-py|absl-py>=2.1.0|g" \
"s|protobuf==3.20.3|protobuf>=3.19.0|g" \
"s|tensorflow-datasets|tensorflow-datasets>=4.8.0|g" \
"s|grain-nightly|grain|g" \
"s|tensorboardx|tensorboardx>=2.6.2|g" \
"s|sentencepiece==0.1.97|sentencepiece>=0.2|g" \
; do sed -i "${pattern}" ${SRC_PATH_MAXTEXT}/requirements.txt; done for requirement in \
"tensorflow-metadata>=1.15.0" \
"seqio@git+https://github.com/google/seqio.git" \
; do echo "${requirement}" >> ${SRC_PATH_MAXTEXT}/requirements.txt done EOF # buildkit
01/22/2025 8:53 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.