Skip to content

Fix DeepSeek 671B recipe to read from quantized checkpoints #90

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Fix DeepSeek 671B recipe to read from quantized checkpoints #90

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
099cfc5
Fix DeepSeek 671B recipe to read from quantized checkpoints
RajeshThallam Sep 17, 2025
e21e564
For DeepSeek inference pin JAX/MaxText to specific commits
RajeshThallam Sep 19, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion inference/trillium/JetStream-Maxtext/DeepSeek-R1-671B/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -363,7 +363,7 @@ helm install -f values.yaml \
--set clusterName=$CLUSTER_NAME \
--set job.jax_tpu_image.repository=${ARTIFACT_REGISTRY}/${JETSTREAM_MAXTEXT_IMAGE} \
--set job.jax_tpu_image.tag=${JETSTREAM_MAXTEXT_VERSION} \
--set maxtext_config.load_parameters_path=${GCS_CKPT_PATH_UNSCANNED}/0/items \
--set maxtext_config.load_parameters_path=${GCS_CKPT_PATH_QUANTIZED} \
jetstream-pathways \
$RECIPE_ROOT/serve-model
```
Expand Down
9 changes: 6 additions & 3 deletions inference/trillium/JetStream-Maxtext/DeepSeek-R1-671B/docker/Dockerfile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,8 +44,11 @@ ENV XCLOUD_ENVIRONMENT=GCP
# Install JetStream and MaxText

RUN git clone https://github.com/AI-Hypercomputer/JetStream.git && \
git clone https://github.com/AI-Hypercomputer/maxtext.git && \
git clone https://github.com/google/aqt.git
cd JetStream && git checkout 29329e8 && cd .. && \
git clone https://github.com/AI-Hypercomputer/maxtext.git && \
cd maxtext && git checkout 4f63ce61 && cd .. && \
git clone https://github.com/google/aqt.git && \
cd aqt && git checkout 28504f5 && cd ..

RUN cd /maxtext && bash setup.sh && pip install torch --index-url https://download.pytorch.org/whl/cpu

Expand All @@ -56,4 +59,4 @@ RUN cd /JetStream && pip install -e .
RUN apt -y update && apt-get -y install python3-dev && apt-get -y install build-essential
RUN cp -r /aqt/aqt/* /usr/local/lib/python3.10/dist-packages/aqt/

ENTRYPOINT [ "/bin/bash" ]
ENTRYPOINT [ "/bin/bash" ]