Gahow Wang
445e491123
third_party/vllm/ now tracked in git for direct patch management.
Based on vLLM v0.18.1 release with one patch applied:
vllm/v1/core/sched/scheduler.py:
Replace fatal assert with graceful skip when KV transfer callback
arrives for an already-aborted request during PD disaggregated serving.
Future vLLM modifications should be made directly in third_party/vllm/
and committed normally. The patches/ directory is kept as documentation
of what changed from upstream.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
41 lines
1.3 KiB
YAML
41 lines
1.3 KiB
YAML
{{- if and .Values.extraInit .Values.extraInit.modelDownload.enabled }}
|
|
apiVersion: batch/v1
|
|
kind: Job
|
|
metadata:
|
|
name: "{{ .Release.Name }}-init-vllm"
|
|
namespace: {{ .Release.Namespace }}
|
|
spec:
|
|
ttlSecondsAfterFinished: 100
|
|
template:
|
|
metadata:
|
|
name: init-vllm
|
|
spec:
|
|
containers:
|
|
- name: job-download-model
|
|
image: {{ .Values.extraInit.modelDownload.image.repository }}:{{ .Values.extraInit.modelDownload.image.tag }}
|
|
imagePullPolicy: {{ .Values.extraInit.modelDownload.image.pullPolicy }}
|
|
command: {{ .Values.extraInit.modelDownload.downloadJob.command | toJson }}
|
|
args:
|
|
{{- toYaml .Values.extraInit.modelDownload.downloadJob.args | nindent 8 }}
|
|
env:
|
|
{{- if .Values.extraInit.modelDownload.downloadJob.env }}
|
|
{{- toYaml .Values.extraInit.modelDownload.downloadJob.env | nindent 8 }}
|
|
{{- else }}
|
|
{{- include "chart.extraInitEnv" . | nindent 8 }}
|
|
{{- end }}
|
|
volumeMounts:
|
|
- name: {{ .Release.Name }}-storage
|
|
mountPath: /data
|
|
resources:
|
|
requests:
|
|
cpu: 200m
|
|
memory: 1Gi
|
|
limits:
|
|
cpu: 500m
|
|
memory: 2Gi
|
|
restartPolicy: OnFailure
|
|
volumes:
|
|
- name: {{ .Release.Name }}-storage
|
|
persistentVolumeClaim:
|
|
claimName: "{{ .Release.Name }}-storage-claim"
|
|
{{- end }} |