discopat.nn_training.torch_detection_utils package

Submodules

discopat.nn_training.torch_detection_utils.coco_eval module

class discopat.nn_training.torch_detection_utils.coco_eval.CocoEvaluator(coco_gt, iou_types)[source]

Bases: object

accumulate()[source]
prepare(predictions, iou_type)[source]
prepare_for_coco_detection(predictions)[source]
prepare_for_coco_keypoint(predictions)[source]
prepare_for_coco_segmentation(predictions)[source]
summarize()[source]
synchronize_between_processes()[source]
update(predictions)[source]
discopat.nn_training.torch_detection_utils.coco_eval.convert_to_xywh(boxes)[source]
discopat.nn_training.torch_detection_utils.coco_eval.create_common_coco_eval(coco_eval, img_ids, eval_imgs)[source]
discopat.nn_training.torch_detection_utils.coco_eval.evaluate(imgs)[source]
discopat.nn_training.torch_detection_utils.coco_eval.merge(img_ids, eval_imgs)[source]

discopat.nn_training.torch_detection_utils.coco_utils module

discopat.nn_training.torch_detection_utils.coco_utils.convert_to_coco_api(ds)[source]
discopat.nn_training.torch_detection_utils.coco_utils.get_coco_api_from_dataset(dataset)[source]

discopat.nn_training.torch_detection_utils.engine module

discopat.nn_training.torch_detection_utils.engine.train_one_epoch(model, optimizer, data_loader, device, epoch, print_freq, scaler=None)[source]

discopat.nn_training.torch_detection_utils.utils module

class discopat.nn_training.torch_detection_utils.utils.MetricLogger(delimiter='\t')[source]

Bases: object

add_meter(name, meter)[source]
log_every(iterable, print_freq, header=None)[source]
synchronize_between_processes()[source]
update(**kwargs)[source]
class discopat.nn_training.torch_detection_utils.utils.SmoothedValue(window_size=20, fmt=None)[source]

Bases: object

Track a series of values and provide access to smoothed values over a window or the global series average.

property avg
property global_avg
property max
property median
synchronize_between_processes()[source]

Warning: does not synchronize the deque!.

update(value, n=1)[source]
property value
discopat.nn_training.torch_detection_utils.utils.all_gather(data)[source]

Run all_gather on arbitrary picklable data (not necessarily tensors).

Parameters:

data – any picklable object

Returns:

list of data gathered from each rank

Return type:

list[data]

discopat.nn_training.torch_detection_utils.utils.collate_fn(batch)[source]
discopat.nn_training.torch_detection_utils.utils.get_rank()[source]
discopat.nn_training.torch_detection_utils.utils.get_world_size()[source]
discopat.nn_training.torch_detection_utils.utils.is_dist_avail_and_initialized()[source]
discopat.nn_training.torch_detection_utils.utils.is_main_process()[source]
discopat.nn_training.torch_detection_utils.utils.mkdir(path)[source]
discopat.nn_training.torch_detection_utils.utils.reduce_dict(input_dict, average=True)[source]

Reduce the values in the dictionary from all processes so that all processes.

Parameters:

  • input_dict (dict) – all the values will be reduced

  • average (bool) – whether to do average or sum have the averaged results.

Returns a dict with the same fields as input_dict, after reduction.

discopat.nn_training.torch_detection_utils.utils.setup_for_distributed(is_master)[source]

Disable printing when not in master process.

Module contents