guild_run: Launch a guild run

In guildai: Track Machine Learning Experiments

Launch a guild run

Description

Launch a guild run

Usage

guild_run(
  opspec = "train.R",
  flags = NULL,
  ...,
  echo = TRUE,
  as_job = getOption("guildai.run_as_job", TRUE)
)

Arguments

opspec	typically a path to an R script, but could be any string that guild recognizes as a valid operation.
flags	flag values for the run(s). A named list or vector like c(noise = .3, dropout = .4). Lists with vectors of length greater than 1 are automatically expanded into a grid of combinations for a batch of runs. For example, list(noise = c(.2, .3), dropout = c(.4, .5) expands to a batch of 4 runs. A dataframe of flags for a batch of runs, one row per run. A scalar string like "noise=.3 dropout=.4". Any flags specification accepted at the terminal is valid here as well.
...	Arguments passed on to guild_run_cli label Set a label for the run. tag Associate TAG with run. May be used multiple times. comment Comment associated with the run. run_dir Use alternative run directory DIR. Cannot be used with stage. stage (bool) Stage an operation. start Start a staged run or restart an existing run. Cannot be used with proto or run_dir. restart Start a staged run or restart an existing run. Cannot be used with proto or run_dir. proto Use the operation, flags and source code from RUN. Flags may be added or redefined in this operation. Cannot be used with restart. force_sourcecode (bool) Use working source code when restart or proto is specified. Ignored otherwise. gpus Limit availabe GPUs to DEVICES, a comma separated list of device IDs. By default all GPUs are available. Cannot beused with no_gpus. no_gpus (bool) Disable GPUs for run. Cannot be used with gpus. batch_label Label to use for batch runs. Ignored for non-batch runs. batch_tag Associate TAG with batch. Ignored for non-batch runs. May be used multiple times. batch_comment Comment associated with batch. optimizer Optimize the run using the specified algorithm. See Optimizing Runs for more information. optimize (bool) Optimize the run using the default optimizer. minimize Column to minimize when running with an optimizer. See help for compare command for details specifying a column. May not be used with maximize. maximize Column to maximize when running with an optimizer. See help for compare command for details specifying a column. May not be used with minimize. opt_flag Flag for OPTIMIZER. May be used multiple times. max_trials Maximum number of trials to run in batch operations. Default is optimizer specific. If optimizer is not specified, default is 20. trials Maximum number of trials to run in batch operations. Default is optimizer specific. If optimizer is not specified, default is 20. stage_trials (bool) For batch operations, stage trials without running them. remote Run the operation remotely. force_flags (bool) Accept all flag assignments, even for undefined or invalid values. force_deps (bool) Continue even when a required resource is not resolved. stop_after Stop operation after N minutes. fail_on_trial_error (bool) Stop batch operations when a trial exits with an error. needed (bool) Run only if there is not an available matching run. A matching run is of the same operation with the same flag values that is not stopped due to an error. background (bool) Run operation in background. pidfile Run operation in background, writing the background process ID to PIDFILE. no_wait (bool) Don't wait for a remote operation to complete. Ignored if run is local. save_trials Saves generated trials to a CSV batch file. See BATCH FILES for more information. keep_run (bool) Keep run even when configured with 'delete-on-success'. keep_batch (bool) Keep batch run rather than delete it on success. dep Include PATH as a dependency. quiet (bool) Do not show output. print_cmd (bool) Show operation command and exit. print_env (bool) Show operation environment and exit. print_trials (bool) Show generated trials and exit. help_model (bool) Show model help and exit. help_op (bool) Show operation help and exit. test_output_scalars Test output scalars on output. Use '-' to read from standard intput. test_sourcecode (bool) Test source code selection. test_flags (bool) Test flag configuration.
echo	whether output from the run is shown in the current R console. Note, this has no effect on whether expressions are echoed in the guild run stdout log. To disable echoing of expression in the run logs, specify #| echo: false in the run script frontmatter.
as_job	Run the operation as an RStudio background job. This is ignored outside of the RStudio IDE.

Value

NULL, invisibly. This function is called for its side effect.

guildai documentation built on March 7, 2023, 5:32 p.m.