Ë çi+ãóT—dZddlmZddlmZddlmZddlZddl m Z Gd„d«Z y) z'Base class used to build new callbacks.é)ÚAny)ÚTensor)Ú OptimizerN)Ú STEP_OUTPUTcó—eZdZdZedefd„«Zededfd„«Zde defd„Z ddd d d edd fd „Z ddd d d edd fd„Z d=d„Z d=d„Zd=d„Zd=d„Zddd d de dedd f d„Zddd d dede dedd f d„Zd=d„Zd=d„Zd=d„Zd=d„Zd=d„Zd=d„Zd=d„Zd=d„Z d>ddd d de ded edd f d!„Z d>ddd d dede ded edd fd"„Z d>ddd d de ded edd f d#„Z d>ddd d dede ded edd fd$„Z d>ddd d de ded edd f d%„Z! d>ddd d de de ded edd fd&„Z"d=d'„Z#d=d(„Z$d=d)„Z%d=d*„Z&d=d+„Z'd=d,„Z(d=d-„Z)d=d.„Z*ddd d d/e+dd fd0„Z,de-ee ffd1„Z.d2e-ee fdd fd3„Z/ddd d d4e-ee fdd fd5„Z0ddd d d4e-ee fdd fd6„Z1ddd d d7e2dd fd8„Z3d=d9„Z4ddd d d:e5dd fd;„Z6ddd d d:e5dd fd<„Z7y )?ÚCallbackzvAbstract base class used to build new callbacks. Subclass this class and override any of the relevant hooks Úreturncó.—|jjS)auIdentifier for the state of the callback. Used to store and retrieve a callback's state from the checkpoint dictionary by ``checkpoint["callbacks"][state_key]``. Implementations of a callback need to provide a unique state key if 1) the callback has state and 2) it is desired to maintain the state of multiple instances of that callback. )Ú __class__Ú __qualname__©Úselfs úy/Volumes/fast/ai/experiments/voice-extract-mac/.venv/lib/python3.12/site-packages/pytorch_lightning/callbacks/callback.pyÚ state_keyzCallback.state_key s€ð~‰~×*Ñ*Ð*ócó—t|«S)z7State key for checkpoints saved prior to version 1.5.0.)Útyper s rÚ_legacy_state_keyzCallback._legacy_state_key+s€ôD‹zÐrÚkwargsc óH—|jj›t|«›S)zÿFormats a set of key-value pairs into a state key string with the callback class name prefixed. Useful for defining a :attr:`state_key`. Args: **kwargs: A set of key-value pairs. Must be serializable to :class:`str`. )r r Úrepr)rrs rÚ_generate_state_keyzCallback._generate_state_key0s"€ð—.‘.×-Ñ-Ð.¬t°F«|¨nÐ=Ð=rÚtrainerú pl.TrainerÚ pl_moduleúpl.LightningModuleÚstageNcó—y)z9Called when fit, validate, test, predict, or tune begins.N©©rrrrs rÚsetupzCallback.setup:órcó—y)z7Called when fit, validate, test, predict, or tune ends.Nrr s rÚteardownzCallback.teardown=r"rcó—y)zCalled when fit begins.Nr©rrrs rÚ on_fit_startzCallback.on_fit_start@r"rcó—y)zCalled when fit ends.Nrr&s rÚ on_fit_endzCallback.on_fit_endCr"rcó—y)z/Called when the validation sanity check starts.Nrr&s rÚon_sanity_check_startzCallback.on_sanity_check_startFr"rcó—y)z-Called when the validation sanity check ends.Nrr&s rÚon_sanity_check_endzCallback.on_sanity_check_endIr"rÚbatchÚ batch_idxcó—y)z#Called when the train batch begins.Nr)rrrr.r/s rÚon_train_batch_startzCallback.on_train_batch_startLr"rÚoutputscó—y)záCalled when the train batch ends. Note: The value ``outputs["loss"]`` here will be the normalized value w.r.t ``accumulate_grad_batches`` of the loss returned from ``training_step``. Nr)rrrr2r.r/s rÚon_train_batch_endzCallback.on_train_batch_endQr"rcó—y)z#Called when the train epoch begins.Nrr&s rÚon_train_epoch_startzCallback.on_train_epoch_start\r"rcó—y)a+Called when the train epoch ends. To access all batch outputs at the end of the epoch, you can cache step outputs as an attribute of the :class:`pytorch_lightning.core.LightningModule` and access them in this hook: .. code-block:: python class MyLightningModule(L.LightningModule): def __init__(self): super().__init__() self.training_step_outputs = [] def training_step(self): loss = ... self.training_step_outputs.append(loss) return loss class MyCallback(L.Callback): def on_train_epoch_end(self, trainer, pl_module): # do something with all training_step outputs, for example: epoch_mean = torch.stack(pl_module.training_step_outputs).mean() pl_module.log("training_epoch_mean", epoch_mean) # free up the memory pl_module.training_step_outputs.clear() Nrr&s rÚon_train_epoch_endzCallback.on_train_epoch_end_r"rcó—y)z!Called when the val epoch begins.Nrr&s rÚon_validation_epoch_startz"Callback.on_validation_epoch_start|r"rcó—y)zCalled when the val epoch ends.Nrr&s rÚon_validation_epoch_endz Callback.on_validation_epoch_endr"rcó—y)z"Called when the test epoch begins.Nrr&s rÚon_test_epoch_startzCallback.on_test_epoch_start‚r"rcó—y)z Called when the test epoch ends.Nrr&s rÚon_test_epoch_endzCallback.on_test_epoch_end…r"rcó—y)z%Called when the predict epoch begins.Nrr&s rÚon_predict_epoch_startzCallback.on_predict_epoch_startˆr"rcó—y)z#Called when the predict epoch ends.Nrr&s rÚon_predict_epoch_endzCallback.on_predict_epoch_end‹r"rÚdataloader_idxcó—y)z(Called when the validation batch begins.Nr©rrrr.r/rEs rÚon_validation_batch_startz"Callback.on_validation_batch_startŽr"rcó—y)z&Called when the validation batch ends.Nr©rrrr2r.r/rEs rÚon_validation_batch_endz Callback.on_validation_batch_end˜r"rcó—y)z"Called when the test batch begins.NrrGs rÚon_test_batch_startzCallback.on_test_batch_start£r"rcó—y)z Called when the test batch ends.NrrJs rÚon_test_batch_endzCallback.on_test_batch_end­r"rcó—y)z%Called when the predict batch begins.NrrGs rÚon_predict_batch_startzCallback.on_predict_batch_start¸r"rcó—y)z#Called when the predict batch ends.NrrJs rÚon_predict_batch_endzCallback.on_predict_batch_endÂr"rcó—y)zCalled when the train begins.Nrr&s rÚon_train_startzCallback.on_train_startÍr"rcó—y)zCalled when the train ends.Nrr&s rÚ on_train_endzCallback.on_train_endÐr"rcó—y)z'Called when the validation loop begins.Nrr&s rÚon_validation_startzCallback.on_validation_startÓr"rcó—y)z%Called when the validation loop ends.Nrr&s rÚon_validation_endzCallback.on_validation_endÖr"rcó—y)zCalled when the test begins.Nrr&s rÚ on_test_startzCallback.on_test_startÙr"rcó—y)zCalled when the test ends.Nrr&s rÚ on_test_endzCallback.on_test_endÜr"rcó—y)zCalled when the predict begins.Nrr&s rÚon_predict_startzCallback.on_predict_startßr"rcó—y)zCalled when predict ends.Nrr&s rÚon_predict_endzCallback.on_predict_endâr"rÚ exceptioncó—y)zACalled when any trainer execution is interrupted by an exception.Nr)rrrrds rÚ on_exceptionzCallback.on_exceptionår"rcó—iS)z¡Called when saving a checkpoint, implement to generate callback's ``state_dict``. Returns: A dictionary containing callback state. rr s rÚ state_dictzCallback.state_dictès €ðˆ rrhcó—y)zÅCalled when loading a checkpoint, implement to reload callback state given callback's ``state_dict``. Args: state_dict: the callback state returned by ``state_dict``. Nr)rrhs rÚload_state_dictzCallback.load_state_dictñs€ð rÚ checkpointcó—y)aCalled when saving a checkpoint to give you a chance to store anything else you might want to save. Args: trainer: the current :class:`~pytorch_lightning.trainer.trainer.Trainer` instance. pl_module: the current :class:`~pytorch_lightning.core.LightningModule` instance. checkpoint: the checkpoint dictionary that will be saved. Nr©rrrrks rÚon_save_checkpointzCallback.on_save_checkpointúr"rcó—y)aiCalled when loading a model checkpoint, use to reload state. Args: trainer: the current :class:`~pytorch_lightning.trainer.trainer.Trainer` instance. pl_module: the current :class:`~pytorch_lightning.core.LightningModule` instance. checkpoint: the full checkpoint dictionary that got loaded by the Trainer. Nrrms rÚon_load_checkpointzCallback.on_load_checkpointr"rÚlosscó—y)z"Called before ``loss.backward()``.Nr)rrrrqs rÚon_before_backwardzCallback.on_before_backwardr"rcó—y)zCCalled after ``loss.backward()`` and before optimizers are stepped.Nrr&s rÚon_after_backwardzCallback.on_after_backwardr"rÚ optimizercó—y)z#Called before ``optimizer.step()``.Nr©rrrrvs rÚon_before_optimizer_stepz!Callback.on_before_optimizer_stepr"rcó—y)z(Called before ``optimizer.zero_grad()``.Nrrxs rÚon_before_zero_gradzCallback.on_before_zero_gradr"r)rrrrr N)r)8Ú__name__Ú __module__r Ú__doc__ÚpropertyÚstrrrrrrr!r$r'r)r+r-Úintr1rr4r6r8r:r<r>r@rBrDrHrKrMrOrQrSrUrWrYr[r]r_rarcÚ BaseExceptionrfÚdictrhrjrnrprrsrurryr{rrrrrsP„ñð ð+˜3ò+óð+ðð 4¨ Ñ#3òóðð>¨Cð>°Có>ðH˜\ðHÐ6JðHÐSVðHÐ[_óHðF  ðFÐ9MðFÐVYðFÐ^bóFó&ó$ó>ó<ð2Ø#ð2Ø0Dð2ØMPð2Ø]`ð2à ó2ð  Ø#ð  Ø0Dð  ØOZð  Øcfð  Øsvð  à ó  ó2ó ó:0ó.ó1ó/ó4ó2ð ñ 7àð7ð(ð7ðð 7ð ð 7ð ð 7ð ó7ð" ñ 5àð 5ð(ð 5ðð 5ð ð 5ð ð 5ðð 5ð ó 5ð" ñ 1àð1ð(ð1ðð 1ð ð 1ð ð 1ð ó1ð" ñ /àð /ð(ð /ðð /ð ð /ð ð /ðð /ð ó /ð" ñ 4àð4ð(ð4ðð 4ð ð 4ð ð 4ð ó4ð" ñ 2àð 2ð(ð 2ðð 2ð ð 2ð ð 2ðð 2ð ó 2ó,ó*ó6ó4ó+ó)ó.ó(ðP LðPÐ=QðPÐ^kðPÐptóPð˜D  c ™Nóð ¨$¨s°C¨x©.ð ¸Tó ð  Ø#ð  Ø0Dð  ØRVÐWZÐ\_ÐW_ÑR`ð  à ó  ð  Ø#ð  Ø0Dð  ØRVÐWZÐ\_ÐW_ÑR`ð  à ó  ð1¨,ð1ÐCWð1Ð_eð1Ðjnó1óRð2Ø#ð2Ø0Dð2ØQZð2à ó2ð 7¨<ð7ÐDXð7Ðenð7Ðswô7rr) r~ÚtypingrÚtorchrÚ torch.optimrÚpytorch_lightningÚplÚ!pytorch_lightning.utilities.typesrrrrrÚrŠs%ðñ/ååÝ!ãÝ9÷E7òE7r