akka optimize training runs
Manage training runs, snapshots and child runs
Synopsis
A training run trains a model using a specified dataset and configuration.
To train and evaluate a model in a single multi-stage workflow, use training pipelines.
Each plan submission starts a new training run. Submitting an identical plan starts a distinct run for that configuration.
A run commits snapshots as it trains. "snapshots" lists them, and "branch" submits a child run from one source snapshot with a budget of additional optimizer updates. "sweep" submits one child run for each point of a parameter grid from a single source snapshot, and can score the source against the resulting fan. "branches" lists direct child runs, "tree" displays descendants across all generations, and "ancestry" traces a run back to its root.
Examples
# Submit a plan and wait for the run to finish akka-optimize training runs start -f plan.json --wait # List active and incomplete runs akka-optimize training runs list --phase open # Follow an existing run and exit with a non-zero status if it fails akka-optimize training runs wait 7d496e82c3a15f04 --exit-status # Display details for a run in JSON format akka-optimize training runs get 7d496e82c3a15f04 -o json # List the snapshots a run committed akka-optimize training runs snapshots 7d496e82c3a15f04
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 runs ancestry - Display the ancestor chain leading to a run
-
akka optimize training runs annotate - Update annotations and configuration name for a training run
-
akka optimize training runs branch - Submit a child run from a source snapshot
-
akka optimize training runs branches - List child runs created from a parent run
-
akka optimize training runs cancel - Cancel an active training run
-
akka optimize training runs engine-config - Display the configuration a training run’s engine was given
-
akka optimize training runs get - Display details for a training run
-
akka optimize training runs grafana - Print a run’s Grafana dashboard URL
-
akka optimize training runs hparams - Display the hyperparameters a training run recorded
-
akka optimize training runs list - List training runs
-
akka optimize training runs metrics - Display step-by-step metrics recorded by a training run
-
akka optimize training runs pause - Pause a run at a selected checkpoint
-
akka optimize training runs resume - Resume a paused training run
-
akka optimize training runs rollouts - Display the rollouts a reinforcement run scored at one step
-
akka optimize training runs snapshots - List a run’s committed snapshots
-
akka optimize training runs start - Start a training run from a plan file
-
akka optimize training runs sweep - Submit one child run for each point of a parameter grid
-
akka optimize training runs tensorboard - Open a run’s training curves in TensorBoard
-
akka optimize training runs tree - Display a tree of runs branched from a root run
-
akka optimize training runs wait - Wait for a training run to finish