Reorganized for new stable/waypoint versioning design.
This commit is contained in:
15
vllm/versions/0/manifest.yaml
Normal file
15
vllm/versions/0/manifest.yaml
Normal file
@@ -0,0 +1,15 @@
|
||||
version: 0.5.4-1
|
||||
requires: []
|
||||
defaultConfig:
|
||||
namespace: llm
|
||||
model: Qwen/Qwen2.5-7B-Instruct
|
||||
maxModelLen: 8192
|
||||
gpuMemoryUtilization: 0.9
|
||||
gpuProduct: RTX 4090
|
||||
cpuRequest: '4'
|
||||
cpuLimit: '8'
|
||||
memoryRequest: 16Gi
|
||||
memoryLimit: 24Gi
|
||||
gpuCount: 1
|
||||
domain: vllm.{{ .cloud.domain }}
|
||||
defaultSecrets: []
|
||||
Reference in New Issue
Block a user