mirror of
https://github.com/nold360/hive-apps
synced 2024-12-22 21:51:20 +00:00
feat(localai): runtimeclass nvidia
This commit is contained in:
parent
d9a93a6fb9
commit
6dba741b68
2 changed files with 30 additions and 19 deletions
|
@ -3,9 +3,12 @@ config:
|
||||||
|
|
||||||
apps:
|
apps:
|
||||||
- name: localai
|
- name: localai
|
||||||
repoURL: https://go-skynet.github.io/helm-charts
|
# repoURL: https://go-skynet.github.io/helm-charts
|
||||||
chart: local-ai
|
# chart: local-ai
|
||||||
targetRevision: 3.1.0
|
# targetRevision: 3.1.0
|
||||||
|
repoURL: https://github.com/nold360/localai-charts.git
|
||||||
|
path: charts/local-ai
|
||||||
|
targetRevision: feat/runtimeclass
|
||||||
|
|
||||||
- name: anythingllm
|
- name: anythingllm
|
||||||
repo: bjw-s
|
repo: bjw-s
|
||||||
|
|
|
@ -1,9 +1,14 @@
|
||||||
replicaCount: 1
|
replicaCount: 1
|
||||||
|
|
||||||
deployment:
|
deployment:
|
||||||
image: quay.io/go-skynet/local-ai:master-ffmpeg-core
|
image:
|
||||||
|
repository: quay.io/go-skynet/local-ai
|
||||||
|
tag: v2.8.0-cublas-cuda12
|
||||||
|
pullPolicy: Always
|
||||||
|
runtimeClassName: nvidia
|
||||||
|
|
||||||
env:
|
env:
|
||||||
threads: 16
|
# threads: 16
|
||||||
context_size: 2048
|
context_size: 2048
|
||||||
DEBUG: "true"
|
DEBUG: "true"
|
||||||
|
|
||||||
|
@ -54,9 +59,6 @@ deployment:
|
||||||
prompt_templates:
|
prompt_templates:
|
||||||
# To use cloud provided (eg AWS) image, provide it like: 1234356789.dkr.ecr.us-REGION-X.amazonaws.com/busybox
|
# To use cloud provided (eg AWS) image, provide it like: 1234356789.dkr.ecr.us-REGION-X.amazonaws.com/busybox
|
||||||
image: busybox
|
image: busybox
|
||||||
pullPolicy: Always
|
|
||||||
imagePullSecrets: []
|
|
||||||
# - name: secret-names
|
|
||||||
|
|
||||||
resources:
|
resources:
|
||||||
requests:
|
requests:
|
||||||
|
@ -80,17 +82,25 @@ models:
|
||||||
|
|
||||||
# The list of URLs to download models from
|
# The list of URLs to download models from
|
||||||
# Note: the name of the file will be the name of the loaded model
|
# Note: the name of the file will be the name of the loaded model
|
||||||
list:
|
list: []
|
||||||
- url: "https://gpt4all.io/models/ggml-gpt4all-j.bin"
|
# - url: "https://gpt4all.io/models/ggml-gpt4all-j.bin"
|
||||||
# basicAuth: base64EncodedCredentials
|
# basicAuth: base64EncodedCredentials
|
||||||
|
|
||||||
persistence:
|
persistence:
|
||||||
pvc:
|
models:
|
||||||
enabled: true
|
enabled: true
|
||||||
|
annotations: {}
|
||||||
|
storageClass: ssd
|
||||||
|
accessModes: ReadWriteOnce
|
||||||
size: 100Gi
|
size: 100Gi
|
||||||
accessModes:
|
globalMount: /models
|
||||||
- ReadWriteOnce
|
output:
|
||||||
storageClass: "ssd"
|
enabled: false
|
||||||
|
annotations: {}
|
||||||
|
storageClass: ssd
|
||||||
|
accessModes: ReadWriteOnce
|
||||||
|
size: 100Gi
|
||||||
|
globalMount: /tmp/generated
|
||||||
|
|
||||||
service:
|
service:
|
||||||
type: ClusterIP
|
type: ClusterIP
|
||||||
|
@ -115,5 +125,3 @@ ingress:
|
||||||
hosts:
|
hosts:
|
||||||
- ai.dc
|
- ai.dc
|
||||||
|
|
||||||
image:
|
|
||||||
pullPolicy: IfNotPresent
|
|
||||||
|
|
Loading…
Reference in a new issue