Index

A

accelerator (xpm_torch.configuration.FabricConfiguration attribute) add_metric() (xpm_torch.trainers.context.TrainerContext method)

after (xpm_torch.module.ReadmeSection attribute) assemble_readme_sections() (in module xpm_torch.module)

B

batch_size (xpm_torch.trainers.LossTrainer attribute) (xpm_torch.trainers.validation.TrainerValidationLoss attribute) batcher (xpm_torch.trainers.LossTrainer attribute) (xpm_torch.trainers.validation.TrainerValidationLoss attribute)

before (xpm_torch.module.ReadmeSection attribute) bestpath (xpm_torch.trainers.validation.TrainerValidationLoss attribute)

C

check_hf_cache() (in module xpm_torch.huggingface) checkpoint_interval (xpm_torch.learner.Learner attribute)

checkpointspath (xpm_torch.learner.Learner attribute) content (xpm_torch.module.ReadmeSection attribute) copy() (xpm_torch.trainers.context.TrainerContext method)

D

data (xpm_torch.parameters.PrefixRenamer attribute) (xpm_torch.trainers.validation.TrainerValidationLoss attribute) default (xpm_torch.parameters.SubParametersIterator attribute)

default_name (xpm_torch.actions.ExportAction attribute) devices (xpm_torch.configuration.FabricConfiguration attribute) download_huggingface_model() (in module xpm_torch.huggingface)

E

early_stop (xpm_torch.trainers.validation.TrainerValidationLoss attribute) epoch (xpm_torch.learner.CheckpointSettings attribute) (xpm_torch.trainers.context.TrainState attribute)

eps (xpm_torch.optim.Adam attribute) (xpm_torch.optim.AdamW attribute) excludes (xpm_torch.optim.RegexParameterFilter attribute)

F

fabric_config (xpm_torch.learner.Learner attribute)

filter (xpm_torch.optim.ParameterOptimizer attribute)

G

get_hf_config() (in module xpm_torch.huggingface)

get_manageable_modules() (xpm_torch.module.ModuleContainer method)

H

hooks (xpm_torch.learner.Learner attribute) (xpm_torch.trainers.LossTrainer attribute) (xpm_torch.trainers.multiple.MultipleTrainer attribute) (xpm_torch.trainers.Trainer attribute)

I

id (xpm_torch.learner.LearnerListener attribute) (xpm_torch.trainers.validation.TrainerValidationLoss attribute) includes (xpm_torch.optim.RegexParameterFilter attribute)

info (xpm_torch.trainers.validation.TrainerValidationLoss attribute) iterator (xpm_torch.parameters.InverseParametersIterator attribute) (xpm_torch.parameters.SubParametersIterator attribute)

K

key (xpm_torch.module.ReadmeSection attribute) (xpm_torch.validation.ValidationSettings attribute)

L

listener (xpm_torch.validation.ValidationSettings attribute) listeners (xpm_torch.learner.Learner attribute) load() (xpm_torch.trainers.context.TrainState method) load_bestcheckpoint() (xpm_torch.trainers.context.TrainerContext method) loader (xpm_torch.actions.ExportAction attribute)

logpath (xpm_torch.learner.Learner attribute) lr (xpm_torch.optim.Adafactor attribute) (xpm_torch.optim.Adam attribute) (xpm_torch.optim.AdamW attribute) (xpm_torch.optim.SGD attribute)

M

mapper (xpm_torch.parameters.PartialModuleLoader attribute) margin (xpm_torch.losses.pairwise.HingeLoss attribute) max_epochs (xpm_torch.learner.Learner attribute) max_norm (xpm_torch.optim.GradientClippingHook attribute) metrics (xpm_torch.trainers.context.TrainerContext attribute) min_factor (xpm_torch.schedulers.LinearWithWarmup attribute) model (xpm_torch.learner.Learner attribute) (xpm_torch.parameters.PrefixRenamer attribute) (xpm_torch.parameters.RegexParametersIterator attribute) (xpm_torch.parameters.SubParametersIterator attribute) (xpm_torch.trainers.LossTrainer attribute) (xpm_torch.trainers.multiple.MultipleTrainer attribute) (xpm_torch.trainers.Trainer attribute)

models (xpm_torch.results.TrainingResults attribute) module xpm_torch.huggingface module (xpm_torch.optim.ParameterOptimizer attribute) ModuleContainer (class in xpm_torch.module)

N

name (xpm_torch.optim.GradientLogHook attribute) negative_regex (xpm_torch.parameters.RegexParametersIterator attribute) num_cycles (xpm_torch.schedulers.CosineWithWarmup attribute)

num_nodes (xpm_torch.configuration.FabricConfiguration attribute) num_warmup_steps (xpm_torch.schedulers.CosineWithWarmup attribute) (xpm_torch.schedulers.LinearWithWarmup attribute) num_workers (xpm_torch.trainers.LossTrainer attribute)

O

optimizer (xpm_torch.optim.ParameterOptimizer attribute)

optimizers (xpm_torch.learner.Learner attribute)

P

path (xpm_torch.module.SimpleModuleLoader attribute) (xpm_torch.parameters.PartialModuleLoader attribute) (xpm_torch.parameters.SubModuleLoader attribute)

precision (xpm_torch.configuration.FabricConfiguration attribute) prepare_hf_model() (in module xpm_torch.huggingface) pretrained_loader() (xpm_torch.huggingface.TorchHFHub class method)

R

random (xpm_torch.learner.Learner attribute) ReadmeSection (class in xpm_torch.module)

regex (xpm_torch.parameters.RegexParametersIterator attribute) relative_step (xpm_torch.optim.Adafactor attribute)

S

sampler (xpm_torch.trainers.LossTrainer attribute) save() (xpm_torch.trainers.context.TrainState method) saved_value (xpm_torch.parameters.SubModuleLoader attribute) scheduler (xpm_torch.optim.ParameterOptimizer attribute) seed (xpm_torch.base.Random attribute) selector (xpm_torch.parameters.PartialModuleLoader attribute) (xpm_torch.parameters.SubModuleLoader attribute) (xpm_torch.trainers.hooks.LayerFreezer attribute)

settings (xpm_torch.module.ModuleLoader attribute) (xpm_torch.module.SimpleModuleLoader attribute) setup_with_fabric() (xpm_torch.module.ModuleContainer method) source (xpm_torch.trainers.hooks.LayerSharer attribute) step (xpm_torch.trainers.context.TrainState property) steps (xpm_torch.trainers.context.TrainState attribute) steps_per_epoch (xpm_torch.learner.Learner attribute) strategy (xpm_torch.configuration.FabricConfiguration attribute) sub_modules (xpm_torch.module.ModuleList attribute)

T

target (xpm_torch.trainers.hooks.LayerSharer attribute) tb_logs (xpm_torch.results.TrainingResults attribute) torch_fp32_precision (xpm_torch.configuration.FabricConfiguration attribute) TorchHFHub (class in xpm_torch.huggingface)

trainer (xpm_torch.learner.Learner attribute) (xpm_torch.trainers.validation.TrainerValidationLoss attribute) TrainerContext (class in xpm_torch.trainers.context) trainers (xpm_torch.trainers.multiple.MultipleTrainer attribute) TrainState (class in xpm_torch.trainers.context)

V

validation_interval (xpm_torch.trainers.validation.TrainerValidationLoss attribute) value (xpm_torch.module.ModuleLoader attribute) (xpm_torch.module.SimpleModuleLoader attribute) (xpm_torch.parameters.PartialModuleLoader attribute) (xpm_torch.parameters.SubModuleLoader attribute)

version (xpm_torch.optim.GradientClippingHook attribute)

W

warmup (xpm_torch.trainers.validation.TrainerValidationLoss attribute) weight (xpm_torch.losses.batchwise.BatchwiseLoss attribute) (xpm_torch.losses.batchwise.CrossEntropyLoss attribute) (xpm_torch.losses.batchwise.SoftmaxCrossEntropy attribute) (xpm_torch.losses.pairwise.CrossEntropyLoss attribute) (xpm_torch.losses.pairwise.HingeLoss attribute) (xpm_torch.losses.pairwise.PairwiseLoss attribute) (xpm_torch.losses.pairwise.PointwiseCrossEntropyLoss attribute)

weight_decay (xpm_torch.optim.Adafactor attribute) (xpm_torch.optim.Adam attribute) (xpm_torch.optim.AdamW attribute) (xpm_torch.optim.SGD attribute) writer (xpm_torch.trainers.context.TrainerContext property)

X

xpm_torch.huggingface module