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

Config cleanup #8

Merged
merged 1 commit into from
Jan 17, 2024
Merged

Config cleanup #8

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
18 changes: 9 additions & 9 deletions configs/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -63,7 +63,7 @@ For list of all nodes, see [nodes](../luxonis_train/nodes/README.md).
| Key | Type | Default value | Description |
| ----------------------- | -------------------- | ------------- | ------------------------------------------------------------------------------------------------------------------------------------------- |
| name | str | | name of the node |
| override_name | str | None | custom name for the node |
| alias | str | None | custom name for the node |
| params | dict | {} | parameters for the node |
| inputs | list | \[\] | list of input nodes for this node, if empty, the node is understood to be an input node of the model |
| freezing.active | bool | False | whether to freeze the modules so the weights are not updated |
Expand All @@ -73,12 +73,12 @@ For list of all nodes, see [nodes](../luxonis_train/nodes/README.md).

Modules that are attached to a node. This include losses, metrics and visualziers.

| Key | Type | Default value | Description |
| ------------- | ---- | ------------- | ------------------------------------------- |
| name | str | | name of the module |
| attached_to | str | | Name of the node the module is attached to. |
| override_name | str | None | custom name for the module |
| params | dict | {} | parameters of the module |
| Key | Type | Default value | Description |
| ----------- | ---- | ------------- | ------------------------------------------- |
| name | str | | name of the module |
| attached_to | str | | Name of the node the module is attached to. |
| alias | str | None | custom name for the module |
| params | dict | {} | parameters of the module |

#### Losses

Expand Down Expand Up @@ -128,9 +128,9 @@ To store and load the data we use LuxonisDataset and LuxonisLoader. For specific

| Key | Type | Default value | Description |
| -------------- | ---------------------------------------- | ------------------- | ---------------------------------------------- |
| dataset_name | str \| None | None | name of the dataset |
| name | str \| None | None | name of the dataset |
| id | str \| None | None | id of the dataset |
| team_id | str \| None | None | team under which you can find all datasets |
| dataset_id | str \| None | None | id of the dataset |
| bucket_type | Literal\["intenal", "external"\] | internal | type of underlying storage |
| bucket_storage | Literal\["local", "s3", "gcc", "azure"\] | BucketStorage.LOCAL | underlying object storage for a bucket |
| train_view | str | train | view to use for training |
Expand Down
2 changes: 1 addition & 1 deletion configs/classification_model.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ model:
include_plot: True

dataset:
dataset_name: cifar10_test
name: cifar10_test

trainer:
preprocessing:
Expand Down
2 changes: 1 addition & 1 deletion configs/coco_model.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -96,7 +96,7 @@ tracker:
is_mlflow: False

dataset:
dataset_name: coco_test
name: coco_test
train_view: train
val_view: val
test_view: test
Expand Down
2 changes: 1 addition & 1 deletion configs/detection_model.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ model:
use_neck: True

dataset:
dataset_name: coco_test
name: coco_test

trainer:
preprocessing:
Expand Down
2 changes: 1 addition & 1 deletion configs/example_export.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ model:
task: binary

dataset:
dataset_name: coco_test
name: coco_test

trainer:
preprocessing:
Expand Down
2 changes: 1 addition & 1 deletion configs/example_tuning.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@ model:
task: binary

dataset:
dataset_name: coco_test
name: coco_test

trainer:
preprocessing:
Expand Down
2 changes: 1 addition & 1 deletion configs/keypoint_bbox_model.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@ model:
name: KeypointDetectionModel

dataset:
dataset_name: coco_test
name: coco_test

trainer:
preprocessing:
Expand Down
2 changes: 1 addition & 1 deletion configs/segmentation_model.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@ model:
task: binary

dataset:
dataset_name: coco_test
name: coco_test

trainer:
preprocessing:
Expand Down
4 changes: 2 additions & 2 deletions luxonis_train/__main__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -126,9 +126,9 @@ def inspect(
image_size = cfg.trainer.preprocessing.train_image_size

dataset = LuxonisDataset(
dataset_name=cfg.dataset.dataset_name,
dataset_name=cfg.dataset.name,
team_id=cfg.dataset.team_id,
dataset_id=cfg.dataset.dataset_id,
dataset_id=cfg.dataset.id,
bucket_type=cfg.dataset.bucket_type,
bucket_storage=cfg.dataset.bucket_storage,
)
Expand Down
2 changes: 1 addition & 1 deletion luxonis_train/callbacks/export_on_train_end.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,7 @@ def on_train_end(self, trainer: pl.Trainer, pl_module: pl.LightningModule) -> No
cfg: Config = pl_module.cfg
cfg.model.weights = best_model_path
if self.upload_to_mlflow:
if pl_module.cfg.tracker.is_mlflow:
if cfg.tracker.is_mlflow:
tracker = cast(LuxonisTrackerPL, trainer.logger)
new_upload_directory = f"mlflow://{tracker.project_id}/{tracker.run_id}"
cfg.exporter.upload_directory = new_upload_directory
Expand Down
3 changes: 2 additions & 1 deletion luxonis_train/callbacks/metadata_logger.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
import pkg_resources
import yaml

from luxonis_train.utils.config import Config
from luxonis_train.utils.registry import CALLBACKS


Expand All @@ -23,7 +24,7 @@ def __init__(self, hyperparams: list[str]):
self.hyperparams = hyperparams

def on_fit_start(self, trainer: pl.Trainer, pl_module: pl.LightningModule) -> None:
cfg = pl_module.cfg
cfg: Config = pl_module.cfg

hparams = {key: cfg.get(key) for key in self.hyperparams}

Expand Down
28 changes: 15 additions & 13 deletions luxonis_train/callbacks/test_on_train_end.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,7 @@
from luxonis_ml.data import LuxonisDataset, ValAugmentations
from torch.utils.data import DataLoader

from luxonis_train.utils.config import Config
from luxonis_train.utils.loaders import LuxonisLoaderTorch, collate_fn
from luxonis_train.utils.registry import CALLBACKS

Expand All @@ -11,31 +12,32 @@ class TestOnTrainEnd(pl.Callback):
"""Callback to perform a test run at the end of the training."""

def on_train_end(self, trainer: pl.Trainer, pl_module: pl.LightningModule) -> None:
cfg: Config = pl_module.cfg

dataset = LuxonisDataset(
dataset_name=pl_module.cfg.dataset.dataset_name,
team_id=pl_module.cfg.dataset.team_id,
dataset_id=pl_module.cfg.dataset.dataset_id,
bucket_type=pl_module.cfg.dataset.bucket_type,
bucket_storage=pl_module.cfg.dataset.bucket_storage,
dataset_name=cfg.dataset.name,
team_id=cfg.dataset.team_id,
dataset_id=cfg.dataset.id,
bucket_type=cfg.dataset.bucket_type,
bucket_storage=cfg.dataset.bucket_storage,
)

loader_test = LuxonisLoaderTorch(
dataset,
view=pl_module.cfg.dataset.test_view,
view=cfg.dataset.test_view,
augmentations=ValAugmentations(
image_size=pl_module.cfg.trainer.preprocessing.train_image_size,
image_size=cfg.trainer.preprocessing.train_image_size,
augmentations=[
i.model_dump()
for i in pl_module.cfg.trainer.preprocessing.augmentations
i.model_dump() for i in cfg.trainer.preprocessing.augmentations
],
train_rgb=pl_module.cfg.trainer.preprocessing.train_rgb,
keep_aspect_ratio=pl_module.cfg.trainer.preprocessing.keep_aspect_ratio,
train_rgb=cfg.trainer.preprocessing.train_rgb,
keep_aspect_ratio=cfg.trainer.preprocessing.keep_aspect_ratio,
),
)
pytorch_loader_test = DataLoader(
loader_test,
batch_size=pl_module.cfg.trainer.batch_size,
num_workers=pl_module.cfg.trainer.num_workers,
batch_size=cfg.trainer.batch_size,
num_workers=cfg.trainer.num_workers,
collate_fn=collate_fn,
)
trainer.test(pl_module, pytorch_loader_test)
4 changes: 2 additions & 2 deletions luxonis_train/core/core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -122,9 +122,9 @@ def __init__(
callbacks=LuxonisProgressBar() if self.cfg.use_rich_text else None,
)
self.dataset = LuxonisDataset(
dataset_name=self.cfg.dataset.dataset_name,
dataset_name=self.cfg.dataset.name,
team_id=self.cfg.dataset.team_id,
dataset_id=self.cfg.dataset.dataset_id,
dataset_id=self.cfg.dataset.id,
bucket_type=self.cfg.dataset.bucket_type,
bucket_storage=self.cfg.dataset.bucket_storage,
)
Expand Down
17 changes: 10 additions & 7 deletions luxonis_train/core/tuner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,19 +26,22 @@ def __init__(self, cfg: str | dict, args: list[str] | tuple[str, ...] | None):
used for config overriding.
"""
super().__init__(cfg, args)
if self.cfg.tuner is None:
raise ValueError("You have to specify the `tuner` section in config.")
self.tune_cfg = self.cfg.tuner

def tune(self) -> None:
"""Runs Optuna tunning of hyperparameters."""

pruner = (
optuna.pruners.MedianPruner()
if self.cfg.tuner.use_pruner
if self.tune_cfg.use_pruner
else optuna.pruners.NopPruner()
)

storage = None
if self.cfg.tuner.storage.active:
if self.cfg.tuner.storage.storage_type == "local":
if self.tune_cfg.storage.active:
if self.tune_cfg.storage.storage_type == "local":
storage = "sqlite:///study_local.db"
else:
storage = "postgresql://{}:{}@{}:{}/{}".format(
Expand All @@ -50,7 +53,7 @@ def tune(self) -> None:
)

study = optuna.create_study(
study_name=self.cfg.tuner.study_name,
study_name=self.tune_cfg.study_name,
storage=storage,
direction="minimize",
pruner=pruner,
Expand All @@ -59,8 +62,8 @@ def tune(self) -> None:

study.optimize(
self._objective,
n_trials=self.cfg.tuner.n_trials,
timeout=self.cfg.tuner.timeout,
n_trials=self.tune_cfg.n_trials,
timeout=self.tune_cfg.timeout,
)

def _objective(self, trial: optuna.trial.Trial) -> float:
Expand Down Expand Up @@ -128,7 +131,7 @@ def _objective(self, trial: optuna.trial.Trial) -> float:

def _get_trial_params(self, trial: optuna.trial.Trial) -> dict[str, Any]:
"""Get trial params based on specified config."""
cfg_tuner = self.cfg.tuner.params
cfg_tuner = self.tune_cfg.params
new_params = {}
for key, value in cfg_tuner.items():
key_info = key.split("_")
Expand Down
4 changes: 2 additions & 2 deletions luxonis_train/models/luxonis_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -142,7 +142,7 @@ def __init__(
for node_cfg in self.cfg.model.nodes:
node_name = node_cfg.name
Node = BaseNode.REGISTRY.get(node_name)
node_name = node_cfg.override_name or node_name
node_name = node_cfg.alias or node_name
if node_cfg.freezing.active:
epochs = self.cfg.trainer.epochs
if node_cfg.freezing.unfreeze_after is None:
Expand Down Expand Up @@ -714,7 +714,7 @@ def _init_attached_module(
storage: Mapping[str, Mapping[str, BaseAttachedModule]],
) -> tuple[str, str]:
Module = registry.get(cfg.name)
module_name = cfg.override_name or cfg.name
module_name = cfg.alias or cfg.name
node_name = cfg.attached_to
module = Module(**cfg.params, node=self.nodes[node_name])
storage[node_name][module_name] = module # type: ignore
Expand Down
12 changes: 6 additions & 6 deletions luxonis_train/models/predefined_models/classification_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,13 +26,13 @@ def nodes(self) -> list[ModelNodeConfig]:
return [
ModelNodeConfig(
name=self.backbone,
override_name="classification_backbone",
alias="classification_backbone",
freezing=self.backbone_params.pop("freezing", {}),
params=self.backbone_params,
),
ModelNodeConfig(
name="ClassificationHead",
override_name="classification_head",
alias="classification_head",
inputs=["classification_backbone"],
freezing=self.head_params.pop("freezing", {}),
params=self.head_params,
Expand All @@ -44,7 +44,7 @@ def losses(self) -> list[LossModuleConfig]:
return [
LossModuleConfig(
name="CrossEntropyLoss",
override_name="classification_loss",
alias="classification_loss",
attached_to="classification_head",
params=self.loss_params,
weight=1.0,
Expand All @@ -56,20 +56,20 @@ def metrics(self) -> list[MetricModuleConfig]:
return [
MetricModuleConfig(
name="F1Score",
override_name="classification_f1_score",
alias="classification_f1_score",
is_main_metric=True,
attached_to="classification_head",
params={"task": self.task},
),
MetricModuleConfig(
name="Accuracy",
override_name="classification_accuracy",
alias="classification_accuracy",
attached_to="classification_head",
params={"task": self.task},
),
MetricModuleConfig(
name="Recall",
override_name="classification_recall",
alias="classification_recall",
attached_to="classification_head",
params={"task": self.task},
),
Expand Down
12 changes: 6 additions & 6 deletions luxonis_train/models/predefined_models/detection_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,7 @@ def nodes(self) -> list[ModelNodeConfig]:
nodes = [
ModelNodeConfig(
name="EfficientRep",
override_name="detection_backbone",
alias="detection_backbone",
freezing=self.backbone_params.pop("freezing", {}),
params=self.backbone_params,
),
Expand All @@ -34,7 +34,7 @@ def nodes(self) -> list[ModelNodeConfig]:
nodes.append(
ModelNodeConfig(
name="RepPANNeck",
override_name="detection_neck",
alias="detection_neck",
inputs=["detection_backbone"],
freezing=self.neck_params.pop("freezing", {}),
params=self.neck_params,
Expand All @@ -44,7 +44,7 @@ def nodes(self) -> list[ModelNodeConfig]:
nodes.append(
ModelNodeConfig(
name="EfficientBBoxHead",
override_name="detection_head",
alias="detection_head",
freezing=self.head_params.pop("freezing", {}),
inputs=["detection_neck"] if self.use_neck else ["detection_backbone"],
params=self.head_params,
Expand All @@ -57,7 +57,7 @@ def losses(self) -> list[LossModuleConfig]:
return [
LossModuleConfig(
name="AdaptiveDetectionLoss",
override_name="detection_loss",
alias="detection_loss",
attached_to="detection_head",
params=self.loss_params,
weight=1.0,
Expand All @@ -69,7 +69,7 @@ def metrics(self) -> list[MetricModuleConfig]:
return [
MetricModuleConfig(
name="MeanAveragePrecision",
override_name="detection_map",
alias="detection_map",
attached_to="detection_head",
is_main_metric=True,
),
Expand All @@ -80,7 +80,7 @@ def visualizers(self) -> list[AttachedModuleConfig]:
return [
AttachedModuleConfig(
name="BBoxVisualizer",
override_name="detection_visualizer",
alias="detection_visualizer",
attached_to="detection_head",
params=self.visualizer_params,
)
Expand Down
Loading