# Copyright (C) 2024 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""Converter for v1 config."""
from __future__ import annotations
import argparse
import json
from copy import deepcopy
from pathlib import Path
from typing import Any
from warnings import warn
from jsonargparse import ArgumentParser, Namespace
from otx.core.config.data import SamplerConfig, SubsetConfig, TileConfig, UnlabeledDataConfig
from otx.core.data.module import OTXDataModule
from otx.core.model.base import OTXModel
from otx.core.types import PathLike
from otx.core.types.task import OTXTaskType
from otx.engine import Engine
from otx.engine.utils.auto_configurator import AutoConfigurator
TEMPLATE_ID_DICT = {
# MULTI_CLASS_CLS
"Custom_Image_Classification_DeiT-Tiny": {
"task": OTXTaskType.MULTI_CLASS_CLS,
"model_name": "deit_tiny",
},
"Custom_Image_Classification_EfficinetNet-B0": {
"task": OTXTaskType.MULTI_CLASS_CLS,
"model_name": "efficientnet_b0",
},
"Custom_Image_Classification_EfficientNet-V2-S": {
"task": OTXTaskType.MULTI_CLASS_CLS,
"model_name": "efficientnet_v2",
},
"Custom_Image_Classification_MobileNet-V3-large-1x": {
"task": OTXTaskType.MULTI_CLASS_CLS,
"model_name": "mobilenet_v3_large",
},
"Custom_Image_Classification_EfficientNet-B3": {
"task": OTXTaskType.MULTI_CLASS_CLS,
"model_name": "tv_efficientnet_b3",
},
"Custom_Image_Classification_EfficientNet-V2-L": {
"task": OTXTaskType.MULTI_CLASS_CLS,
"model_name": "tv_efficientnet_v2_l",
},
"Custom_Image_Classification_MobileNet-V3-small": {
"task": OTXTaskType.MULTI_CLASS_CLS,
"model_name": "tv_mobilenet_v3_small",
},
# DETECTION
"Custom_Object_Detection_Gen3_ATSS": {
"task": OTXTaskType.DETECTION,
"model_name": "atss_mobilenetv2",
},
"Object_Detection_ResNeXt101_ATSS": {
"task": OTXTaskType.DETECTION,
"model_name": "atss_resnext101",
},
"Custom_Object_Detection_Gen3_SSD": {
"task": OTXTaskType.DETECTION,
"model_name": "ssd_mobilenetv2",
},
"Object_Detection_YOLOX_X": {
"task": OTXTaskType.DETECTION,
"model_name": "yolox_x",
},
"Object_Detection_YOLOX_L": {
"task": OTXTaskType.DETECTION,
"model_name": "yolox_l",
},
"Object_Detection_YOLOX_S": {
"task": OTXTaskType.DETECTION,
"model_name": "yolox_s",
},
"Custom_Object_Detection_YOLOX": {
"task": OTXTaskType.DETECTION,
"model_name": "yolox_tiny",
},
"Object_Detection_RTDetr_18": {
"task": OTXTaskType.DETECTION,
"model_name": "rtdetr_18",
},
"Object_Detection_RTDetr_50": {
"task": OTXTaskType.DETECTION,
"model_name": "rtdetr_50",
},
"Object_Detection_RTDetr_101": {
"task": OTXTaskType.DETECTION,
"model_name": "rtdetr_101",
},
"Object_Detection_RTMDet_tiny": {
"task": OTXTaskType.DETECTION,
"model_name": "rtmdet_tiny",
},
# INSTANCE_SEGMENTATION
"Custom_Counting_Instance_Segmentation_MaskRCNN_ResNet50": {
"task": OTXTaskType.INSTANCE_SEGMENTATION,
"model_name": "maskrcnn_r50",
},
"Custom_Counting_Instance_Segmentation_MaskRCNN_SwinT_FP16": {
"task": OTXTaskType.INSTANCE_SEGMENTATION,
"model_name": "maskrcnn_swint",
},
"Custom_Counting_Instance_Segmentation_MaskRCNN_EfficientNetB2B": {
"task": OTXTaskType.INSTANCE_SEGMENTATION,
"model_name": "maskrcnn_efficientnetb2b",
},
"Custom_Instance_Segmentation_RTMDet_tiny": {
"task": OTXTaskType.INSTANCE_SEGMENTATION,
"model_name": "rtmdet_inst_tiny",
},
# ROTATED_DETECTION
"Custom_Rotated_Detection_via_Instance_Segmentation_MaskRCNN_ResNet50": {
"task": OTXTaskType.ROTATED_DETECTION,
"model_name": "maskrcnn_r50",
},
"Custom_Rotated_Detection_via_Instance_Segmentation_MaskRCNN_EfficientNetB2B": {
"task": OTXTaskType.ROTATED_DETECTION,
"model_name": "maskrcnn_efficientnetb2b",
},
# SEMANTIC_SEGMENTATION
"Custom_Semantic_Segmentation_Lite-HRNet-18-mod2_OCR": {
"task": OTXTaskType.SEMANTIC_SEGMENTATION,
"model_name": "litehrnet_18",
},
"Custom_Semantic_Segmentation_Lite-HRNet-18_OCR": {
"task": OTXTaskType.SEMANTIC_SEGMENTATION,
"model_name": "litehrnet_18",
},
"Custom_Semantic_Segmentation_Lite-HRNet-s-mod2_OCR": {
"task": OTXTaskType.SEMANTIC_SEGMENTATION,
"model_name": "litehrnet_s",
},
"Custom_Semantic_Segmentation_Lite-HRNet-x-mod3_OCR": {
"task": OTXTaskType.SEMANTIC_SEGMENTATION,
"model_name": "litehrnet_x",
},
"Custom_Semantic_Segmentation_SegNext_t": {
"task": OTXTaskType.SEMANTIC_SEGMENTATION,
"model_name": "segnext_t",
},
"Custom_Semantic_Segmentation_SegNext_s": {
"task": OTXTaskType.SEMANTIC_SEGMENTATION,
"model_name": "segnext_s",
},
"Custom_Semantic_Segmentation_SegNext_B": {
"task": OTXTaskType.SEMANTIC_SEGMENTATION,
"model_name": "segnext_b",
},
"Custom_Semantic_Segmentation_DINOV2_S": {
"task": OTXTaskType.SEMANTIC_SEGMENTATION,
"model_name": "dino_v2",
},
# ANOMALY
"ote_anomaly_padim": {
"task": OTXTaskType.ANOMALY,
"model_name": "padim",
},
"ote_anomaly_stfpm": {
"task": OTXTaskType.ANOMALY,
"model_name": "stfpm",
},
# ANOMALY CLASSIFICATION
"ote_anomaly_classification_padim": {
"task": OTXTaskType.ANOMALY_CLASSIFICATION,
"model_name": "padim",
},
"ote_anomaly_classification_stfpm": {
"task": OTXTaskType.ANOMALY_CLASSIFICATION,
"model_name": "stfpm",
},
# ANOMALY_DETECTION
"ote_anomaly_detection_padim": {
"task": OTXTaskType.ANOMALY_DETECTION,
"model_name": "padim",
},
"ote_anomaly_detection_stfpm": {
"task": OTXTaskType.ANOMALY_DETECTION,
"model_name": "stfpm",
},
# ANOMALY_SEGMENTATION
"ote_anomaly_segmentation_padim": {
"task": OTXTaskType.ANOMALY_SEGMENTATION,
"model_name": "padim",
},
"ote_anomaly_segmentation_stfpm": {
"task": OTXTaskType.ANOMALY_SEGMENTATION,
"model_name": "stfpm",
},
# KEYPOINT_DETECTION
"Keypoint_Detection_RTMPose_Tiny": {
"task": OTXTaskType.KEYPOINT_DETECTION,
"model_name": "rtmpose_tiny",
},
}
[docs]
class ConfigConverter:
"""Convert ModelTemplate for OTX v1 to OTX v2 recipe dictionary.
This class is used to convert ModelTemplate for OTX v1 to OTX v2 recipe dictionary.
Example:
The following examples show how to use the Converter class.
We expect a config file with ModelTemplate information in json form.
Convert template.json to dictionary::
converter = ConfigConverter()
config = converter.convert("template.json")
Instantiate an object from the configuration dictionary::
engine, train_kwargs = converter.instantiate(
config=config,
work_dir="otx-workspace",
data_root="tests/assets/car_tree_bug",
)
Train the model::
engine.train(**train_kwargs)
"""
[docs]
@staticmethod
def convert(config_path: str, task: OTXTaskType | None = None) -> dict:
"""Convert a configuration file to a default configuration dictionary.
Args:
config_path (str): The path to the configuration file.
task (OTXTaskType | None): Value to override the task.
Returns:
dict: The default configuration dictionary.
"""
with Path(config_path).open() as f:
template_config = json.load(f)
hyperparameters = template_config["hyperparameters"]
param_dict = ConfigConverter._get_params(hyperparameters)
task_info = TEMPLATE_ID_DICT[template_config["model_template_id"]]
if param_dict.get("enable_tiling", None) and not task_info["model_name"].endswith("_tile"):
task_info["model_name"] += "_tile"
# classification task type can't be deducted from template name, try to extract from config
if "sub_task_type" in template_config and "_CLS" in task_info["task"]:
task_info["task"] = template_config["sub_task_type"]
if task is not None:
task_info["task"] = task
default_config = ConfigConverter._get_default_config(task_info)
ConfigConverter._update_params(default_config, param_dict)
if (hpo_time_ratio := template_config.get("hpo_parameters", {}).get("hpo_time_ratio")) is not None:
default_config["hpo_config.expected_time_ratio"] = hpo_time_ratio
ConfigConverter._remove_unused_key(default_config)
return default_config
@staticmethod
def _get_default_config(task_info: dict) -> dict:
"""Return default otx conifg for template use."""
return AutoConfigurator(**task_info).config # type: ignore[arg-type]
@staticmethod
def _get_params(hyperparameters: dict) -> dict:
"""Get configuraable parameters from ModelTemplate config hyperparameters field."""
param_dict = {}
for param_name, param_info in hyperparameters.items():
if isinstance(param_info, dict):
if "value" in param_info:
param_dict[param_name] = param_info["value"]
else:
param_dict = param_dict | ConfigConverter._get_params(param_info)
return param_dict
@staticmethod
def _update_params(config: dict, param_dict: dict) -> None: # noqa: C901
"""Update params of OTX recipe from Geit configurable params."""
unused_params = deepcopy(param_dict)
def update_mem_cache_size(param_value: int) -> None:
config["data"]["mem_cache_size"] = f"{int(param_value / 1000000)}MB"
def update_batch_size(param_value: int) -> None:
config["data"]["train_subset"]["batch_size"] = param_value
def update_inference_batch_size(param_value: int) -> None:
config["data"]["val_subset"]["batch_size"] = param_value
config["data"]["test_subset"]["batch_size"] = param_value
def update_learning_rate(param_value: float) -> None:
optimizer = config["model"]["init_args"]["optimizer"]
if isinstance(optimizer, dict) and "init_args" in optimizer:
optimizer["init_args"]["lr"] = param_value
else:
warn("Warning: learning_rate is not updated", stacklevel=1)
def update_learning_rate_warmup_iters(param_value: int) -> None:
scheduler = config["model"]["init_args"]["scheduler"]
if (
isinstance(scheduler, dict)
and "class_path" in scheduler
and scheduler["class_path"] == "otx.core.schedulers.LinearWarmupSchedulerCallable"
):
scheduler["init_args"]["num_warmup_steps"] = param_value
else:
warn("Warning: learning_rate_warmup_iters is not updated", stacklevel=1)
def update_num_iters(param_value: int) -> None:
config["max_epochs"] = param_value
def update_num_workers(param_value: int) -> None:
config["data"]["train_subset"]["num_workers"] = param_value
config["data"]["val_subset"]["num_workers"] = param_value
config["data"]["test_subset"]["num_workers"] = param_value
def update_enable_early_stopping(param_value: bool) -> None:
idx = ConfigConverter._get_callback_idx(
config["callbacks"],
"otx.algo.callbacks.adaptive_early_stopping.EarlyStoppingWithWarmup",
)
if not param_value and idx > -1:
config["callbacks"].pop(idx)
def update_early_stop_patience(param_value: int) -> None:
for callback in config["callbacks"]:
if callback["class_path"] == "otx.algo.callbacks.adaptive_early_stopping.EarlyStoppingWithWarmup":
callback["init_args"]["patience"] = param_value
break
def update_use_adaptive_interval(param_value: bool) -> None:
idx = ConfigConverter._get_callback_idx(
config["callbacks"],
"otx.algo.callbacks.adaptive_train_scheduling.AdaptiveTrainScheduling",
)
if not param_value and idx > -1:
config["callbacks"].pop(idx)
def update_auto_num_workers(param_value: bool) -> None:
config["data"]["auto_num_workers"] = param_value
def update_auto_adapt_batch_size(param_value: str) -> None:
config["adaptive_bs"] = param_value
def update_enable_tiling(param_value: bool) -> None:
config["data"]["tile_config"]["enable_tiler"] = param_value
if param_value:
config["data"]["tile_config"]["enable_adaptive_tiling"] = param_dict["enable_adaptive_params"]
config["data"]["tile_config"]["tile_size"] = (
param_dict["tile_size"],
param_dict["tile_size"],
)
config["data"]["tile_config"]["overlap"] = param_dict["tile_overlap"]
config["data"]["tile_config"]["max_num_instances"] = param_dict["tile_max_number"]
config["data"]["tile_config"]["sampling_ratio"] = param_dict["tile_sampling_ratio"]
config["data"]["tile_config"]["object_tile_ratio"] = param_dict["object_tile_ratio"]
tile_params = [
"enable_adaptive_params",
"tile_size",
"tile_overlap",
"tile_max_number",
"tile_sampling_ratio",
"object_tile_ratio",
]
for tile_param in tile_params:
unused_params.pop(tile_param)
param_update_funcs = {
"mem_cache_size": update_mem_cache_size,
"batch_size": update_batch_size,
"inference_batch_size": update_inference_batch_size,
"learning_rate": update_learning_rate,
"learning_rate_warmup_iters": update_learning_rate_warmup_iters,
"num_iters": update_num_iters,
"num_workers": update_num_workers,
"enable_early_stopping": update_enable_early_stopping,
"early_stop_patience": update_early_stop_patience,
"use_adaptive_interval": update_use_adaptive_interval,
"auto_num_workers": update_auto_num_workers,
"enable_tiling": update_enable_tiling,
"auto_adapt_batch_size": update_auto_adapt_batch_size,
}
for param_name, param_value in param_dict.items():
update_func = param_update_funcs.get(param_name)
if update_func:
update_func(param_value) # type: ignore[operator]
unused_params.pop(param_name)
warn("Warning: These parameters are not updated", stacklevel=1)
for param_name, param_value in unused_params.items():
print(f"\t {param_name}: {param_value}")
@staticmethod
def _get_callback_idx(callbacks: list, name: str) -> int:
"""Return required callbacks index from callback list."""
for idx, callback in enumerate(callbacks):
if callback["class_path"] == name:
return idx
return -1
@staticmethod
def _remove_unused_key(config: dict) -> None:
"""Remove unused keys from the config dictionary.
Args:
config (dict): The configuration dictionary.
"""
config.pop("config") # Remove config key that for CLI
config["data"].pop("__path__") # Remove __path__ key that for CLI overriding
[docs]
@staticmethod
def instantiate(
config: dict,
work_dir: PathLike | None = None,
data_root: PathLike | None = None,
**kwargs,
) -> tuple[Engine, dict[str, Any]]:
"""Instantiate an object from the configuration dictionary.
Args:
config (dict): The configuration dictionary.
work_dir (PathLike): Path to the working directory.
data_root (PathLike): The root directory for data.
Returns:
tuple: A tuple containing the engine and the train kwargs dictionary.
"""
config.update(kwargs)
# Instantiate datamodule
data_config = config.pop("data")
if data_root is not None:
data_config["data_root"] = data_root
train_config = data_config.pop("train_subset")
val_config = data_config.pop("val_subset")
test_config = data_config.pop("test_subset")
unlabeled_config = data_config.pop("unlabeled_subset")
datamodule = OTXDataModule(
train_subset=SubsetConfig(sampler=SamplerConfig(**train_config.pop("sampler", {})), **train_config),
val_subset=SubsetConfig(sampler=SamplerConfig(**val_config.pop("sampler", {})), **val_config),
test_subset=SubsetConfig(sampler=SamplerConfig(**test_config.pop("sampler", {})), **test_config),
unlabeled_subset=UnlabeledDataConfig(**unlabeled_config),
tile_config=TileConfig(**data_config.pop("tile_config", {})),
**data_config,
)
# Update num_classes & Instantiate Model
model_config = config.pop("model")
model_config["init_args"]["label_info"] = datamodule.label_info
model_parser = ArgumentParser()
model_parser.add_subclass_arguments(OTXModel, "model", required=False, fail_untyped=False, skip={"label_info"})
model = model_parser.instantiate_classes(Namespace(model=model_config)).get("model")
if hasattr(model, "tile_config"):
model.tile_config = datamodule.tile_config
# Instantiate Engine
config_work_dir = config.pop("work_dir", config["engine"].pop("work_dir", None))
config["engine"]["work_dir"] = work_dir if work_dir is not None else config_work_dir
engine = Engine(
model=model,
datamodule=datamodule,
**config.pop("engine"),
)
# Instantiate Engine.train Arguments
engine_parser = ArgumentParser()
train_arguments = engine_parser.add_method_arguments(
Engine,
"train",
skip={"accelerator", "devices"},
fail_untyped=False,
)
# Update callbacks & logger dir as engine.work_dir
for callback in config["callbacks"]:
if "init_args" in callback and "dirpath" in callback["init_args"]:
callback["init_args"]["dirpath"] = engine.work_dir
for logger in config["logger"]:
if "save_dir" in logger["init_args"]:
logger["init_args"]["save_dir"] = engine.work_dir
if "log_dir" in logger["init_args"]:
logger["init_args"]["log_dir"] = engine.work_dir
instantiated_kwargs = engine_parser.instantiate_classes(Namespace(**config))
train_kwargs = {k: v for k, v in instantiated_kwargs.items() if k in train_arguments}
return engine, train_kwargs
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("-c", "--config", help="Input ModelTemplate config")
parser.add_argument("-i", "--data_root", help="Input dataset root path")
parser.add_argument("-o", "--work_dir", help="Input work directory path")
args = parser.parse_args()
otx_config = ConfigConverter.convert(config_path=args.config)
engine, train_kwargs = ConfigConverter.instantiate(
config=otx_config,
data_root=args.data_root,
work_dir=args.work_dir,
)
engine.train(**train_kwargs)