akka optimize training

Base models, workloads, runs, snapshots, comparisons, pipelines, evaluations, endpoints and their data

Synopsis

A base model is what everything here trains over, and base-models is what this deployment allows and what you select from. Selecting one registers it and caches its weights; nothing holds a GPU until you ask for work.

A run trains a model from a registered dataset against a workload. An evaluation scores a model against a dataset with a grader. A pipeline executes curate, train and evaluate stages in one submission. An endpoint serves a base model or a candidate at an address so you can try it by hand.

A run commits snapshots as it trains. Snapshot weights can be registered as a model to score, or used as the source of a child run with training runs branch.

A session is a raw conversation corpus, and a curation judges one into a training dataset.

capacity reports what holds the GPU pools, which is what a queued run is behind. info reports which backend the service came up on, which is what to check before a run spends anything.

Options

  -h, --help   help for training

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