akka optimize training base-models
See which base models this deployment allows, and select one to use
Synopsis
The allowed list is what an operator permits and this deployment’s hardware can hold, so a model that is listed is one you can use. A model the hardware cannot hold is not listed at all.
A base model is named by the repository the listing shows. An entry id or the model id a run names resolves too, for a script that stored one.
Listing and selecting hold no GPU. Selecting registers the model and fetches its weights into the cluster’s cache, which is the only wait in this area and is paid once per model per deployment. Runs, evaluations and endpoints may name a model the moment it is selected; work queued behind one that is still preparing waits for the weights rather than being refused.
Examples
# See what you can use akka-optimize training base-models list # Select a model by the repository the listing shows akka-optimize training base-models select Qwen/Qwen3-8B # Watch how far preparing it has got akka-optimize training base-models get Qwen/Qwen3-8B # Or wait for one that is already being prepared akka-optimize training base-models wait Qwen/Qwen3-8B
Options inherited from parent commands
--cache-file string location of cache file (default "~/.akka/cache.yaml")
--config string location of config file (default "~/.akka/config.yaml")
--context string configuration context to use
--disable-prompt Disable all interactive prompts when running akka commands. If input is required, defaults will be used, or an error will be raised.
This is equivalent to setting the environment variable AKKA_DISABLE_PROMPTS to true.
--force Skip the confirmation prompt
--jq string Filter the JSON output with a jq expression
--json string Output JSON with only these fields; --json= lists the fields a command offers
--no-color Print without color, as NO_COLOR does
-o, --output string set output format to one of [text,json,json-compact,go-template=] (default "text")
--page-mode string the mode for paging, either paged, buffered or auto. (default "auto")
-q, --quiet set quiet output (helpful when used as part of a script)
--region string Region to use if the project has more than one region
--timeout duration client command timeout (default 10s)
--trainer-url string Trainer service URL (default http://localhost:9001, or TRAINER_URL)
--url string Optimize service URL (default http://localhost:9010, or OPTIMIZE_URL)
--use-grpc-web use grpc-web when talking to Akka APIs. This is useful when behind corporate firewalls that decrypt traffic but don't support HTTP/2.
--verbose set verbose output
SEE ALSO
-
akka optimize training - Base models, workloads, runs, snapshots, comparisons, pipelines, evaluations, endpoints and their data
-
akka optimize training base-models candidates - List an author’s repositories and what this deployment would make of each
-
akka optimize training base-models deselect - Drop a base model selection
-
akka optimize training base-models get - Display one allowed base model, its selection status, and its admitted runs
-
akka optimize training base-models list - List the base models selected here, or every one this deployment allows
-
akka optimize training base-models select - Select a base model for runs, evaluations, and endpoints
-
akka optimize training base-models wait - Wait for a base model to finish being prepared