Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Clean up unneccesary todos in _get_model_state_from_last_generator_run #2476

Closed
wants to merge 2 commits into from
Closed

Clean up unneccesary todos in _get_model_state_from_last_generator_run #2476

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3a2d2fd
Update gen from multiple nodes to support generating multiple arms pe…
mgarrard May 22, 2024
c97487b
Clean up unneccesary todos in _get_model_state_from_last_generator_ru…
mgarrard May 22, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
42 changes: 30 additions & 12 deletions ax/modelbridge/generation_strategy.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -379,9 +379,8 @@ def gen_with_multiple_nodes(
self,
experiment: Experiment,
data: Optional[Data] = None,
n: int = 1, # Total arms to generate
pending_observations: Optional[Dict[str, List[ObservationFeatures]]] = None,
**kwargs: Any, # TODO: @mgarrard Ensure correct dispatch to nodes
arms_per_node: Optional[Dict[str, int]] = None,
) -> List[GeneratorRun]:
"""Produces a List of GeneratorRuns for a single trial, either ``Trial`` or
``BatchTrial``, and if producing a ``BatchTrial`` allows for multiple
Expand All @@ -398,27 +397,53 @@ def gen_with_multiple_nodes(
data: Optional data to be passed to the underlying model's `gen`, which
is called within this method and actually produces the resulting
generator run. By default, data is all data on the `experiment`.
n: Integer representing how total arms to generate for this trial.
pending_observations: A map from metric name to pending
observations for that metric, used by some models to avoid
resuggesting points that are currently being evaluated.
arms_per_node: A map from node name to the number of arms to generate
from that node. If not provided, the number of arms to generate
from each node defaults to one.

Returns:
A list of ``GeneratorRuns`` for a single trial.
"""
# TODO: @mgarrard merge into gen method, just starting here to derisk
# Validate `arms_per_node` if specified, otherwise construct the default
# behavior with keys being node names and values being 1 to represent
# generating a single GR from each node.
node_names = [node.node_name for node in self._nodes]
if arms_per_node is not None and not all(
node_name in arms_per_node for node_name in node_names
):
raise UserInputError(
f"""
Each node defined in the GenerationStrategy must have
an associated number of arms to generate from that node
defined in `arms_per_node`. {arms_per_node} does not
include all of {node_names}. It may be helpful to double check
the spelling.
"""
)
if arms_per_node is None:
arms_per_node = {node_name: 1 for node_name in node_names}
grs = []
continue_gen_for_trial = True

while continue_gen_for_trial:
next_node_name = self.current_node_name
should_transition, next_node = self._curr.should_transition_to_next_node(
raise_data_required_error=False
)
if should_transition:
assert next_node is not None
next_node_name = next_node
grs.extend(
self._gen_multiple(
experiment=experiment,
num_generator_runs=1,
data=data,
n=n,
n=arms_per_node[next_node_name],
pending_observations=pending_observations,
**kwargs,
)
)
continue_gen_for_trial = self._should_continue_gen_for_trial()
Expand Down Expand Up @@ -862,11 +887,6 @@ def _maybe_transition_to_next_node(

def _get_model_state_from_last_generator_run(self) -> Dict[str, Any]:
lgr = self.last_generator_run
# NOTE: This will not be easily compatible with `GenerationNode`;
# will likely need to find last generator run per model. Not a problem
# for now though as GS only allows `GenerationStep`-s for now.
# Potential solution: store generator runs on `GenerationNode`-s and
# split them per-model there.
model_state_on_lgr = {}
# Need to check if model_spec_to_gen_from is none to account for
# ExternalGenerationNodes which leverage models from outside Ax.
Expand All @@ -885,8 +905,6 @@ def _get_model_state_from_last_generator_run(self) -> Dict[str, Any]:

if grs_equal and lgr._model_state_after_gen:
if self.model or isinstance(model_on_curr, ModelRegistryBase):
# TODO[drfreund]: Consider moving this to `GenerationStep` or
# `GenerationNode`.
model_cls = (
self.model.model.__class__
if self.model is not None
Expand Down
161 changes: 114 additions & 47 deletions ax/modelbridge/tests/test_generation_strategy.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -233,6 +233,53 @@ def setUp(self) -> None:
),
],
)
self.complex_multinode_per_trial_gs = GenerationStrategy(
nodes=[
GenerationNode(
node_name="sobol",
model_specs=[self.sobol_model_spec],
transition_criteria=self.single_running_trial_criterion,
),
GenerationNode(
node_name="gpei",
model_specs=[self.gpei_model_spec],
transition_criteria=[
AutoTransitionAfterGenCriterion(
transition_to="sobol_2",
)
],
),
GenerationNode(
node_name="sobol_2",
model_specs=[self.sobol_model_spec],
transition_criteria=[
AutoTransitionAfterGenCriterion(transition_to="sobol_3")
],
),
GenerationNode(
node_name="sobol_3",
model_specs=[self.sobol_model_spec],
transition_criteria=[
MaxTrials(
threshold=2,
transition_to="sobol_4",
block_transition_if_unmet=True,
only_in_statuses=[TrialStatus.RUNNING],
use_all_trials_in_exp=True,
),
AutoTransitionAfterGenCriterion(
transition_to="gpei",
block_transition_if_unmet=True,
continue_trial_generation=False,
),
],
),
GenerationNode(
node_name="sobol_4",
model_specs=[self.sobol_model_spec],
),
],
)

def tearDown(self) -> None:
self.torch_model_bridge_patcher.stop()
Expand Down Expand Up @@ -1453,6 +1500,70 @@ def test_transition_edges(self) -> None:
},
)

def test_multiple_arms_per_node(self) -> None:
"""Test that a ``GenerationStrategy`` which expects some trials to be composed
of multiple nodes can generate multiple arms per node using `arms_per_node`.
"""
exp = get_branin_experiment()
gs = self.complex_multinode_per_trial_gs
gs.experiment = exp
# first check that arms_per node validation works
arms_per_node = {
"sobol": 3,
"sobol_2": 2,
"sobol_3": 1,
"sobol_4": 4,
}
with self.assertRaisesRegex(UserInputError, "defined in `arms_per_node`"):
gs.gen_with_multiple_nodes(exp, arms_per_node=arms_per_node)

# now we will check that the first trial contains 3 arms, the sconed trial
# contains 6 arms (2 from gpei, 1 from sobol_2, 3 from sobol_3), and all
# remaining trials contain 4 arms
arms_per_node = {
"sobol": 3,
"gpei": 1,
"sobol_2": 2,
"sobol_3": 3,
"sobol_4": 4,
}
# for the first trial, we start on sobol, we generate the trial, but it hasn't
# been run yet, so we remain on sobol
trial0 = exp.new_batch_trial(
generator_runs=gs.gen_with_multiple_nodes(exp, arms_per_node=arms_per_node)
)
self.assertEqual(len(trial0.arms_by_name), 3)
self.assertEqual(trial0.generator_runs[0]._generation_node_name, "sobol")
trial0.run()

# after trial 0 is run, we create a trial with nodes gpei, sobol_2, and sobol_3
# However, the sobol_3 criterion requires that we have two running trials. We
# don't move onto sobol_4 until we have two running trials, instead we reset
# to the last first node in a trial.
for _i in range(0, 2):
trial = exp.new_batch_trial(
generator_runs=gs.gen_with_multiple_nodes(
exp, arms_per_node=arms_per_node
)
)
self.assertEqual(gs.current_node_name, "sobol_3")
self.assertEqual(len(trial.arms_by_name), 6)
self.assertEqual(len(trial.generator_runs), 3)
self.assertEqual(trial.generator_runs[0]._generation_node_name, "gpei")
self.assertEqual(len(trial.generator_runs[0].arms), 1)
self.assertEqual(trial.generator_runs[1]._generation_node_name, "sobol_2")
self.assertEqual(len(trial.generator_runs[1].arms), 2)
self.assertEqual(trial.generator_runs[2]._generation_node_name, "sobol_3")
self.assertEqual(len(trial.generator_runs[2].arms), 3)

# after running the next trial should be made from sobol 4
trial.run()
trial = exp.new_batch_trial(
generator_runs=gs.gen_with_multiple_nodes(exp, arms_per_node=arms_per_node)
)
self.assertEqual(trial.generator_runs[0]._generation_node_name, "sobol_4")
self.assertEqual(len(trial.generator_runs[0].arms), 4)

def test_node_gs_with_auto_transitions(self) -> None:
"""Test that node-based generation strategies which leverage
AutoTransitionAfterGen criterion correctly transition and create trials.
Expand Down Expand Up @@ -1496,6 +1607,7 @@ def test_node_gs_with_auto_transitions(self) -> None:
],
)
exp = get_branin_experiment()
gs.experiment = exp

# for the first trial, we start on sobol, we generate the trial, but it hasn't
# been run yet, so we remain on sobol, after the trial is run, the subsequent
Expand All @@ -1517,53 +1629,8 @@ def test_node_gs_with_auto_transitions(self) -> None:

def test_node_gs_with_auto_transitions_three_phase(self) -> None:
exp = get_branin_experiment()
gs_2 = GenerationStrategy(
nodes=[
GenerationNode(
node_name="sobol",
model_specs=[self.sobol_model_spec],
transition_criteria=self.single_running_trial_criterion,
),
GenerationNode(
node_name="gpei",
model_specs=[self.gpei_model_spec],
transition_criteria=[
AutoTransitionAfterGenCriterion(
transition_to="sobol_2",
)
],
),
GenerationNode(
node_name="sobol_2",
model_specs=[self.sobol_model_spec],
transition_criteria=[
AutoTransitionAfterGenCriterion(transition_to="sobol_3")
],
),
GenerationNode(
node_name="sobol_3",
model_specs=[self.sobol_model_spec],
transition_criteria=[
MaxTrials(
threshold=2,
transition_to="sobol_4",
block_transition_if_unmet=True,
only_in_statuses=[TrialStatus.RUNNING],
use_all_trials_in_exp=True,
),
AutoTransitionAfterGenCriterion(
transition_to="gpei",
block_transition_if_unmet=True,
continue_trial_generation=False,
),
],
),
GenerationNode(
node_name="sobol_4",
model_specs=[self.sobol_model_spec],
),
],
)
gs_2 = self.complex_multinode_per_trial_gs
gs_2.experiment = exp

# for the first trial, we start on sobol, we generate the trial, but it hasn't
# been run yet, so we remain on sobol
Expand Down