Model create
Usage #
aioli model create [-h] [--image IMAGE] [--description DESCRIPTION] [--url URL] [--registry REGISTRY]
[--format FORMAT] [-a ARG] [-e ENV] [--gpu-type GPU_TYPE] [--limits-cpu LIMITS_CPU]
[--limits-memory LIMITS_MEMORY] [--limits-gpu LIMITS_GPU] [--requests-cpu REQUESTS_CPU]
[--requests-memory REQUESTS_MEMORY] [--requests-gpu REQUESTS_GPU]
name
Full Example #
aioli model create NAME --image VALUE --description VALUE --url VALUE --registry VALUE --modelformat VALUE --arg VALUE --env VALUE --gpu-type VALUE --limits-cpu VALUE --limits-memory VALUE --limits-gpu VALUE --requests-cpu VALUE --requests-memory VALUE --requests-gpu VALUE
Options: #
The name of the packaged model. Must begin with a letter, but may contain letters, numbers, and hyphen
–image #
Docker container image servicing the packaged model
–description #
Description of the packaged model
–url #
Reference within the specified registry
–registry #
The name or ID of the packaged model registry
–format, –modelformat #
Model format for downloaded models (bento-archive, openllm, nim, unspecified)
-a, –arg #
Argument to be added to the service command line. If specifying an argument that starts with a ‘-’, use the form –arg=
-e, –env #
Specifies an environment variable & value as name=value, to be passed to the launched container
–gpu-type #
GPU type required
–limits-cpu #
CPU limit
–limits-memory #
Memory limit
–limits-gpu #
GPU limit
–requests-cpu #
CPU request
–requests-memory #
Memory request
–requests-gpu #
GPU request