huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-08 19:04:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

chore(neuron): update to optimum-neuron 0.3.0

Browse Source 
Dependencies were changed accordingly, because Neuron SDK was updated to
v2.24.
This commit is contained in:
Alvaro Moran 2025-07-21 12:51:30 +00:00
parent 24c2bff659
commit b11d663ca0
1 changed files with 11 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
Dockerfile.neuron
View File

@ -5,7 +5,7 @@ RUN mkdir -p /tgi
# Fetch the optimum-neuron sources directly to avoid relying on pypi deployments
FROM alpine AS optimum-neuron
RUN mkdir -p /optimum-neuron
ADD https://github.com/huggingface/optimum-neuron/archive/refs/tags/v0.2.2.tar.gz /optimum-neuron/sources.tar.gz
ADD https://github.com/huggingface/optimum-neuron/archive/refs/tags/v0.3.0.tar.gz /optimum-neuron/sources.tar.gz
RUN tar -C /optimum-neuron -xf /optimum-neuron/sources.tar.gz --strip-components=1
# Build cargo components (adapted from TGI original Dockerfile)
@ -108,10 +108,10 @@ RUN wget -qO - https://apt.repos.neuron.amazonaws.com/GPG-PUB-KEY-AMAZON-AWS-NEU
# Install neuronx packages
RUN apt-get update -y \
&& apt-get install -y --no-install-recommends \
aws-neuronx-dkms=2.20.28.0 \
aws-neuronx-collectives=2.24.59.0-838c7fc8b \
aws-neuronx-runtime-lib=2.24.53.0-f239092cc \
aws-neuronx-tools=2.22.61.0 \
aws-neuronx-dkms=2.22.2.0 \
aws-neuronx-collectives=2.26.43.0-47cc904ea \
aws-neuronx-runtime-lib=2.26.42.0-2ff3b5c7d \
aws-neuronx-tools=2.24.54.0 \
libxml2 \
&& rm -rf /var/lib/apt/lists/* \
&& apt-get clean
@ -120,15 +120,15 @@ ENV PATH="/opt/bin/:/opt/aws/neuron/bin:${PATH}"
# Install manually torch CPU version to avoid pulling CUDA
RUN pip3 install \
torch==2.5.1 \
torchvision==0.20.1 \
torch==2.7.0 \
torchvision==0.22.0 \
--index-url https://download.pytorch.org/whl/cpu
RUN pip3 install \
neuronx-cc==2.17.194.0 \
torch-neuronx==2.5.1.2.6.0 \
neuronx-distributed==0.11.0 \
libneuronxla==2.2.1630.0 \
neuronx-cc==2.19.8089.0+8ab9f450 \
torch-neuronx==2.7.0.2.8.6734+ac864f72 \
neuronx-distributed==0.13.14393+b8569585 \
libneuronxla==2.2.4410.0+835a67fb \
--extra-index-url=https://pip.repos.neuron.amazonaws.com
# Install HuggingFace packages