WIP: Introduce TrialRunner Abstraction #720
Draft
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…nit test for bench (not tested)
…e bulk registration (check for is_warm_up)
Co-authored-by: Brian Kroth <bpkroth@users.noreply.github.com>
… parallel-async-trial-runners
…m in parallel eventually
bpkroth
commented
Mar 19, 2024
| @@ -123,6 +137,7 @@ def get_results_df( | |||
| 'tunable_config_id', | |||
| 'tunable_config_trial_group_id', | |||
| 'status', | |||
| 'trial_runner_id', | |||
bpkroth
commented
Mar 19, 2024
| """Get the running state of the current TrialRunner.""" | ||
| return self._is_running | ||
|
|
||
| def run_trial(self, |
bpkroth
commented
Mar 19, 2024
mlos_bench/mlos_bench/launcher.py
Outdated
| self.root_env_config, TunableGroups(), env_global_config, service=self._parent_service) | ||
| self.trial_runners[trial_runner_id] = TrialRunner(trial_runner_id, env) | ||
| _LOG.info("Init %d trial runners for environments: %s", | ||
| self.trial_runners, list(trial_runner.environment for trial_runner in self.trial_runners)) |
bpkroth
commented
Mar 19, 2024
| }) | ||
| # Rotate which TrialRunner the Trial is assigned to. | ||
| self._current_trial_runner_idx = (self._current_trial_runner_idx + 1) % len(self._trial_runners) |
bpkroth
commented
Mar 19, 2024
mlos_bench/mlos_bench/launcher.py
Outdated
| parser.add_argument( | ||
| '--trial_runners', '--trial-runners', required=False, type=int, default=1, | ||
| help='Number of trial runners to run in parallel. ' | ||
| + 'Individual TrialRunners can be identified in configs with $trial_runner_id.') |
bpkroth
commented
Mar 19, 2024
| TrialRunner | ||
| """ | ||
| if trial.trial_runner_id is None: | ||
| raise ValueError(f"Trial {trial} has no trial_runner_id") |
There was a problem hiding this comment.
- Add fallback support for assigning a Trial to a TrialRunner if one is missing.
motus
reviewed
Mar 21, 2024
Comment on lines
57
to
+58
| # pylint: disable=too-many-statements | ||
| # pylint: disable=too-complex |
There was a problem hiding this comment.
Suggested change
| # pylint: disable=too-many-statements | |
| # pylint: disable=too-complex | |
| # pylint: disable=too-many-statements,too-complex |
|
|
||
| # NOTE: Init tunable values *after* the Environment, but *before* the Optimizer | ||
| self.trial_runners: List[TrialRunner] = [] | ||
| for trial_runner_id in range(0, self.global_config["num_trial_runners"]): |
There was a problem hiding this comment.
Suggested change
| for trial_runner_id in range(0, self.global_config["num_trial_runners"]): | |
| for trial_runner_id in range(self.global_config["num_trial_runners"]): |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is another step in adding support for parallel trial execution #380.
Here we separate out the running of an individual trial to a single class - TrialRunner.
Multiple TrialRunners are instantiated at CLI invocation with the
--num-trial-runnersargument.Each TrialRunner associated with a single copy of the root Environment, and made unique by means of a unique
trial_runner_idvalue that's included in that Environment's global_config.TODO:
In future PRs we will add: