# Copyright (c) Facebook, Inc. and its affiliates.
"""
Registry is central source of truth in MMF. Inspired from Redux's
concept of global store, Registry maintains mappings of various information
to unique keys. Special functions in registry can be used as decorators to
register different kind of classes.
Import the global registry object using
``from mmf.common.registry import registry``
Various decorators for registry different kind of classes with unique keys
- Register a trainer: ``@registry.register_trainer``
- Register a dataset builder: ``@registry.register_builder``
- Register a callback function: ``@registry.register_callback``
- Register a metric: ``@registry.register_metric``
- Register a loss: ``@registry.register_loss``
- Register a fusion technique: ``@registery.register_fusion``
- Register a model: ``@registry.register_model``
- Register a processor: ``@registry.register_processor``
- Register a optimizer: ``@registry.register_optimizer``
- Register a scheduler: ``@registry.register_scheduler``
- Register a encoder: ``@registry.register_encoder``
- Register a decoder: ``@registry.register_decoder``
- Register a transformer backend: ``@registry.register_transformer_backend``
- Register a transformer head: ``@registry.register_transformer_head``
- Register a test reporter: ``@registry.register_test_reporter``
- Register a pl datamodule: ``@registry.register_datamodule``
"""
from mmf.utils.env import setup_imports
[docs]class Registry:
r"""Class for registry object which acts as central source of truth
for MMF
"""
mapping = {
# Mappings of builder name to their respective classes
# Use `registry.register_builder` to register a builder class
# with a specific name
# Further, use the name with the class is registered in the
# command line or configuration to load that specific dataset
"builder_name_mapping": {},
# Similar to the builder_name_mapping above except that this
# one is used to keep a mapping for dataset to its trainer class.
"trainer_name_mapping": {},
"model_name_mapping": {},
"metric_name_mapping": {},
"loss_name_mapping": {},
"pool_name_mapping": {},
"fusion_name_mapping": {},
"optimizer_name_mapping": {},
"scheduler_name_mapping": {},
"processor_name_mapping": {},
"encoder_name_mapping": {},
"decoder_name_mapping": {},
"transformer_backend_name_mapping": {},
"transformer_head_name_mapping": {},
"test_reporter_mapping": {},
"iteration_strategy_name_mapping": {},
"state": {},
"callback_name_mapping": {},
}
[docs] @classmethod
def register_trainer(cls, name):
r"""Register a trainer to registry with key 'name'
Args:
name: Key with which the trainer will be registered.
Usage::
from mmf.common.registry import registry
from mmf.trainers.custom_trainer import CustomTrainer
@registry.register_trainer("custom_trainer")
class CustomTrainer():
...
"""
def wrap(trainer_cls):
cls.mapping["trainer_name_mapping"][name] = trainer_cls
return trainer_cls
return wrap
[docs] @classmethod
def register_builder(cls, name):
r"""Register a dataset builder to registry with key 'name'
Args:
name: Key with which the metric will be registered.
Usage::
from mmf.common.registry import registry
from mmf.datasets.base_dataset_builder import BaseDatasetBuilder
@registry.register_builder("vqa2")
class VQA2Builder(BaseDatasetBuilder):
...
"""
def wrap(builder_cls):
from mmf.datasets.base_dataset_builder import BaseDatasetBuilder
assert issubclass(
builder_cls, BaseDatasetBuilder
), "All builders must inherit BaseDatasetBuilder class"
cls.mapping["builder_name_mapping"][name] = builder_cls
return builder_cls
return wrap
[docs] @classmethod
def register_callback(cls, name):
r"""Register a callback to registry with key 'name'
Args:
name: Key with which the callback will be registered.
Usage::
from mmf.common.registry import registry
from mmf.trainers.callbacks.base import Callback
@registry.register_callback("logistic")
class LogisticCallback(Callback):
...
"""
def wrap(func):
from mmf.trainers.callbacks.base import Callback
assert issubclass(
func, Callback
), "All callbacks must inherit Callback class"
cls.mapping["callback_name_mapping"][name] = func
return func
return wrap
[docs] @classmethod
def register_metric(cls, name):
r"""Register a metric to registry with key 'name'
Args:
name: Key with which the metric will be registered.
Usage::
from mmf.common.registry import registry
from mmf.modules.metrics import BaseMetric
@registry.register_metric("r@1")
class RecallAt1(BaseMetric):
...
"""
def wrap(func):
from mmf.modules.metrics import BaseMetric
assert issubclass(
func, BaseMetric
), "All Metric must inherit BaseMetric class"
cls.mapping["metric_name_mapping"][name] = func
return func
return wrap
[docs] @classmethod
def register_loss(cls, name):
r"""Register a loss to registry with key 'name'
Args:
name: Key with which the loss will be registered.
Usage::
from mmf.common.registry import registry
from torch import nn
@registry.register_task("logit_bce")
class LogitBCE(nn.Module):
...
"""
def wrap(func):
from torch import nn
assert issubclass(
func, nn.Module
), "All loss must inherit torch.nn.Module class"
cls.mapping["loss_name_mapping"][name] = func
return func
return wrap
[docs] @classmethod
def register_pooler(cls, name):
r"""Register a modality pooling method to registry with key 'name'
Args:
name: Key with which the pooling method will be registered.
Usage::
from mmf.common.registry import registry
from torch import nn
@registry.register_pool("average_pool")
class average_pool(nn.Module):
...
"""
def wrap(func):
from torch import nn
assert issubclass(
func, nn.Module
), "All pooling methods must inherit torch.nn.Module class"
cls.mapping["pool_name_mapping"][name] = func
return func
return wrap
[docs] @classmethod
def register_fusion(cls, name):
r"""Register a fusion technique to registry with key 'name'
Args:
name: Key with which the fusion technique will be registered
Usage::
from mmf.common.registry import registry
from torch import nn
@registry.register_fusion("linear_sum")
class LinearSum():
...
"""
def wrap(func):
from torch import nn
assert issubclass(
func, nn.Module
), "All Fusion must inherit torch.nn.Module class"
cls.mapping["fusion_name_mapping"][name] = func
return func
return wrap
[docs] @classmethod
def register_model(cls, name):
r"""Register a model to registry with key 'name'
Args:
name: Key with which the model will be registered.
Usage::
from mmf.common.registry import registry
from mmf.models.base_model import BaseModel
@registry.register_task("pythia")
class Pythia(BaseModel):
...
"""
def wrap(func):
from mmf.models.base_model import BaseModel
assert issubclass(
func, BaseModel
), "All models must inherit BaseModel class"
cls.mapping["model_name_mapping"][name] = func
return func
return wrap
[docs] @classmethod
def register_processor(cls, name):
r"""Register a processor to registry with key 'name'
Args:
name: Key with which the processor will be registered.
Usage::
from mmf.common.registry import registry
from mmf.datasets.processors import BaseProcessor
@registry.register_task("glove")
class GloVe(BaseProcessor):
...
"""
def wrap(func):
from mmf.datasets.processors.processors import BaseProcessor
assert issubclass(
func, BaseProcessor
), "All Processor classes must inherit BaseProcessor class"
cls.mapping["processor_name_mapping"][name] = func
return func
return wrap
@classmethod
def register_optimizer(cls, name):
def wrap(func):
cls.mapping["optimizer_name_mapping"][name] = func
return func
return wrap
@classmethod
def register_scheduler(cls, name):
def wrap(func):
cls.mapping["scheduler_name_mapping"][name] = func
return func
return wrap
@classmethod
def register_transformer_backend(cls, name):
def wrap(func):
cls.mapping["transformer_backend_name_mapping"][name] = func
return func
return wrap
@classmethod
def register_transformer_head(cls, name):
def wrap(func):
cls.mapping["transformer_head_name_mapping"][name] = func
return func
return wrap
@classmethod
def register_test_reporter(cls, name):
def wrap(func):
cls.mapping["test_reporter_mapping"][name] = func
return func
return wrap
[docs] @classmethod
def register_decoder(cls, name):
r"""Register a decoder to registry with key 'name'
Args:
name: Key with which the decoder will be registered.
Usage::
from mmf.common.registry import registry
from mmf.utils.text import TextDecoder
@registry.register_decoder("nucleus_sampling")
class NucleusSampling(TextDecoder):
...
"""
def wrap(decoder_cls):
from mmf.utils.text import TextDecoder
assert issubclass(
decoder_cls, TextDecoder
), "All decoders must inherit TextDecoder class"
cls.mapping["decoder_name_mapping"][name] = decoder_cls
return decoder_cls
return wrap
[docs] @classmethod
def register_encoder(cls, name):
r"""Register a encoder to registry with key 'name'
Args:
name: Key with which the encoder will be registered.
Usage::
from mmf.common.registry import registry
from mmf.modules.encoders import Encoder
@registry.register_encoder("transformer")
class TransformerEncoder(Encoder):
...
"""
def wrap(encoder_cls):
from mmf.modules.encoders import Encoder
assert issubclass(
encoder_cls, Encoder
), "All encoders must inherit Encoder class"
cls.mapping["encoder_name_mapping"][name] = encoder_cls
return encoder_cls
return wrap
[docs] @classmethod
def register_datamodule(cls, name):
r"""Register a datamodule to registry with key 'name'
Args:
name: Key with which the datamodule will be registered.
Usage::
from mmf.common.registry import registry
import pytorch_lightning as pl
@registry.register_datamodule("my_datamodule")
class MyDataModule(pl.LightningDataModule):
...
"""
def wrap(datamodule_cls):
import pytorch_lightning as pl
assert issubclass(
datamodule_cls, pl.LightningDataModule
), "All datamodules must inherit PyTorch Lightning DataModule class"
cls.mapping["builder_name_mapping"][name] = datamodule_cls
return datamodule_cls
return wrap
[docs] @classmethod
def register_iteration_strategy(cls, name):
r"""Register an iteration_strategy to registry with key 'name'
Args:
name: Key with which the iteration_strategy will be registered.
Usage::
from dataclasses import dataclass
from mmf.common.registry import registry
from mmf.datasets.iterators import IterationStrategy
@registry.register_iteration_strategy("my_iteration_strategy")
class MyStrategy(IterationStrategy):
@dataclass
class Config:
name: str = "my_strategy"
def __init__(self, config, dataloader):
...
"""
def wrap(iteration_strategy_cls):
from mmf.datasets.iteration_strategies import IterationStrategy
assert issubclass(
iteration_strategy_cls, IterationStrategy
), "All datamodules must inherit IterationStrategy class"
cls.mapping["iteration_strategy_name_mapping"][
name
] = iteration_strategy_cls
return iteration_strategy_cls
return wrap
[docs] @classmethod
def register(cls, name, obj):
r"""Register an item to registry with key 'name'
Args:
name: Key with which the item will be registered.
Usage::
from mmf.common.registry import registry
registry.register("config", {})
"""
path = name.split(".")
current = cls.mapping["state"]
for part in path[:-1]:
if part not in current:
current[part] = {}
current = current[part]
current[path[-1]] = obj
@classmethod
def get_trainer_class(cls, name):
return cls.mapping["trainer_name_mapping"].get(name, None)
@classmethod
def get_builder_class(cls, name):
return cls.mapping["builder_name_mapping"].get(name, None)
@classmethod
def get_callback_class(cls, name):
return cls.mapping["callback_name_mapping"].get(name, None)
@classmethod
def get_model_class(cls, name):
return cls.mapping["model_name_mapping"].get(name, None)
@classmethod
def get_processor_class(cls, name):
return cls.mapping["processor_name_mapping"].get(name, None)
@classmethod
def get_metric_class(cls, name):
return cls.mapping["metric_name_mapping"].get(name, None)
@classmethod
def get_loss_class(cls, name):
return cls.mapping["loss_name_mapping"].get(name, None)
@classmethod
def get_pool_class(cls, name):
return cls.mapping["pool_name_mapping"].get(name, None)
@classmethod
def get_optimizer_class(cls, name):
return cls.mapping["optimizer_name_mapping"].get(name, None)
@classmethod
def get_scheduler_class(cls, name):
return cls.mapping["scheduler_name_mapping"].get(name, None)
@classmethod
def get_decoder_class(cls, name):
return cls.mapping["decoder_name_mapping"].get(name, None)
@classmethod
def get_encoder_class(cls, name):
return cls.mapping["encoder_name_mapping"].get(name, None)
@classmethod
def get_iteration_strategy_class(cls, name):
return cls.mapping["iteration_strategy_name_mapping"].get(name, None)
@classmethod
def get_transformer_backend_class(cls, name):
return cls.mapping["transformer_backend_name_mapping"].get(name, None)
@classmethod
def get_transformer_head_class(cls, name):
return cls.mapping["transformer_head_name_mapping"].get(name, None)
@classmethod
def get_test_rerporter_class(cls, name):
return cls.mapping["test_reporter_mapping"].get(name, None)
[docs] @classmethod
def get(cls, name, default=None, no_warning=False):
r"""Get an item from registry with key 'name'
Args:
name (string): Key whose value needs to be retrieved.
default: If passed and key is not in registry, default value will
be returned with a warning. Default: None
no_warning (bool): If passed as True, warning when key doesn't exist
will not be generated. Useful for MMF's
internal operations. Default: False
Usage::
from mmf.common.registry import registry
config = registry.get("config")
"""
original_name = name
name = name.split(".")
value = cls.mapping["state"]
for subname in name:
value = value.get(subname, default)
if value is default:
break
if (
"writer" in cls.mapping["state"]
and value == default
and no_warning is False
):
cls.mapping["state"]["writer"].warning(
"Key {} is not present in registry, returning default value "
"of {}".format(original_name, default)
)
return value
[docs] @classmethod
def unregister(cls, name):
r"""Remove an item from registry with key 'name'
Args:
name: Key which needs to be removed.
Usage::
from mmf.common.registry import registry
config = registry.unregister("config")
"""
return cls.mapping["state"].pop(name, None)
registry = Registry()
# Only setup imports in main process, this means registry won't be
# fully available in spawned child processes (such as dataloader processes)
# but instantiated. This is to prevent issues such as
# https://github.com/facebookresearch/mmf/issues/355
if __name__ == "__main__":
setup_imports()
