bstein/titan-iac
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(ai): warm fast chat model

Browse Source
This commit is contained in:
jenkins 2026-06-29 15:06:44 -03:00
parent 6d32504d67
commit 73d161185d
1 changed files with 6 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
services/ai-llm/deployment.yaml
View File

@ -20,7 +20,7 @@ spec:
labels: labels:
app: ollama app: ollama
annotations: annotations:
ai.bstein.dev/model: qwen2.5:14b-instruct-q4_0 ai.bstein.dev/model: qwen2.5-coder:7b-instruct-q4_0,qwen2.5:14b-instruct-q4_0
ai.bstein.dev/gpu: GPU pool (titan-20/21) ai.bstein.dev/gpu: GPU pool (titan-20/21)
ai.bstein.dev/restartedAt: "2026-01-26T12:00:00Z" ai.bstein.dev/restartedAt: "2026-01-26T12:00:00Z"
spec: spec:
@ -53,6 +53,8 @@ spec:
value: /root/.ollama value: /root/.ollama
- name: OLLAMA_MODEL - name: OLLAMA_MODEL
value: qwen2.5:14b-instruct-q4_0 value: qwen2.5:14b-instruct-q4_0
- name: OLLAMA_FAST_MODEL
value: qwen2.5-coder:7b-instruct-q4_0
command: command:
- /bin/sh - /bin/sh
- -c - -c
@ -60,7 +62,9 @@ spec:
set -e set -e
ollama serve >/tmp/ollama.log 2>&1 & ollama serve >/tmp/ollama.log 2>&1 &
sleep 6 sleep 6
ollama pull "${OLLAMA_MODEL}" for model in "${OLLAMA_FAST_MODEL}" "${OLLAMA_MODEL}"; do
ollama pull "${model}"
done
pkill ollama || true pkill ollama || true
volumeMounts: volumeMounts:
- name: models - name: models