A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 09:45 UTC
[neuron] fix authorization issue (#18364)
Signed-off-by: Liangfu Chen <liangfc@amazon.com>
Liangfu Chen committed
May 19, 2025 at 16:30 UTC
d565e0976fb5ffd353727066ac8aa98272e318af
1 file changed
+3
-1
.buildkite/scripts/hardware_ci/run-neuron-test.sh
+3
-1
@@ -11,13 +11,14 @@ container_name="neuron_$(tr -dc A-Za-z0-9 < /dev/urandom | head -c 10; echo)"
11
HF_CACHE="$(realpath ~)/huggingface"
12
mkdir -p "${HF_CACHE}"
13
HF_MOUNT="/root/.cache/huggingface"
14
+HF_TOKEN=$(aws secretsmanager get-secret-value --secret-id "ci/vllm-neuron/hf-token" --region us-west-2 --query 'SecretString' --output text | jq -r .VLLM_NEURON_CI_HF_TOKEN)
15
16
NEURON_COMPILE_CACHE_URL="$(realpath ~)/neuron_compile_cache"
17
mkdir -p "${NEURON_COMPILE_CACHE_URL}"
18
NEURON_COMPILE_CACHE_MOUNT="/root/.cache/neuron_compile_cache"
19
20
# Try building the docker image
20
-aws ecr get-login-password --region us-west-2 | docker login --username AWS --password-stdin 763104351884.dkr.ecr.us-west-2.amazonaws.com
21
+aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws
22
23
# prune old image and containers to save disk space, and only once a day
24
# by using a timestamp file in tmp.
@@ -47,6 +48,7 @@ trap remove_docker_container EXIT
48
docker run --rm -it --device=/dev/neuron0 --network bridge \
49
-v "${HF_CACHE}:${HF_MOUNT}" \
50
-e "HF_HOME=${HF_MOUNT}" \
51
+ -e "HF_TOKEN=${HF_TOKEN}" \
52
-v "${NEURON_COMPILE_CACHE_URL}:${NEURON_COMPILE_CACHE_MOUNT}" \
53
-e "NEURON_COMPILE_CACHE_URL=${NEURON_COMPILE_CACHE_MOUNT}" \
54
--name "${container_name}" \