easycv.runner package

Submodules

easycv.runner.ev_runner module

class easycv.runner.ev_runner.EVRunner(model, batch_processor=None, optimizer=None, work_dir=None, logger=None, meta=None, fp16_enable=False)

Bases: mmcv.runner.epoch_based_runner.EpochBasedRunner

__init__(model, batch_processor=None, optimizer=None, work_dir=None, logger=None, meta=None, fp16_enable=False)

Epoch Runner for easycv, add support for oss IO and file sync.

Parameters

• model (torch.nn.Module) – The model to be run.

• batch_processor (callable) – A callable method that process a data batch. The interface of this method should be batch_processor(model, data, train_mode) -> dict

• optimizer (dict or torch.optim.Optimizer) – It can be either an optimizer (in most cases) or a dict of optimizers (in models that requires more than one optimizer, e.g., GAN).

• work_dir (str, optional) – The working directory to save checkpoints and logs. Defaults to None.

• logger (logging.Logger) – Logger used during training. Defaults to None. (The default value is just for backward compatibility)

• meta (dict | None) – A dict records some import information such as environment info and seed, which will be logged in logger hook. Defaults to None.

• fp16_enable (bool) – if use fp16

run_iter(data_batch, train_mode, **kwargs)

process for each iteration.

Parameters

• data_batch – Batch of dict of data.

• train_model (bool) – If set True, run training step else validation step.

train(data_loader, **kwargs)

Training process for one epoch which will iterate through all training data and call hooks at different stages.

Parameters

data_loader – data loader object for training

val(data_loader, **kwargs)

Validation step which Deprecated, using evaluation hook instead.

save_checkpoint(out_dir, filename_tmpl='epoch_{}.pth', save_optimizer=True, meta=None, create_symlink=True)

Save checkpoint to file.

Parameters

• out_dir – Directory where checkpoint files are to be saved.

• filename_tmpl (str, optional) – Checkpoint filename pattern.

• save_optimizer (bool, optional) – save optimizer state.

• meta (dict, optional) – Metadata to be saved in checkpoint.

current_lr()

Get current learning rates.

Returns

Current learning rates of all

param groups. If the runner has a dict of optimizers, this method will return a dict.

Return type

list[float] | dict[str, list[float]]

load_checkpoint(filename, map_location=device(type='cpu'), strict=False, logger=None)

Load checkpoint from a file or URL.

Parameters

• filename (str) – Accept local filepath, URL, torchvision://xxx, open-mmlab://xxx, oss://xxx. Please refer to docs/source/model_zoo.md for details.

• map_location (str) – Same as torch.load().

• strict (bool) – Whether to allow different params for the model and checkpoint.

• logger (logging.Logger or None) – The logger for error message.

Returns

The loaded checkpoint.

Return type

dict or OrderedDict

resume(checkpoint, resume_optimizer=True, map_location='default')

Resume state dict from checkpoint.

Parameters

• checkpoint – Checkpoint path

• resume_optimizer – Whether to resume optimizer state

• map_location (str) – Same as torch.load().