Functions — github.com/Nota-NetsPresso/netspresso-trainer

hub / github.com/Nota-NetsPresso/netspresso-trainer / functions

Functions1,065 in github.com/Nota-NetsPresso/netspresso-trainer

⨍ Functions 1,065◇ Types & classes 251

↓ 53 callersMethodupdate

(self, val: Union[float, int], n: int = 1)

src/netspresso_trainer/utils/record.py:67

↓ 24 callersMethodupdate

(self, intersection, union)

src/netspresso_trainer/metrics/segmentation/metric.py:40

↓ 23 callersMethod__init__

(self, in_channels: int, out_channels: int, part_channels:

src/netspresso_trainer/models/op/custom.py:1083

↓ 21 callersMethodempty

(cls)

src/netspresso_trainer/utils/protocols.py:10

↓ 20 callersMethodlog

( self, prefix: Literal['training', 'validation', 'evaluation', 'inference'], epoch: O

src/netspresso_trainer/loggers/base.py:139

↓ 17 callersMethodget

(self, name, as_pop=True)

src/netspresso_trainer/utils/record.py:121

↓ 13 callersMethodeval

(self)

src/netspresso_trainer/models/base.py:164

↓ 11 callersMethodstep

(self, epoch: int, metric=None)

src/netspresso_trainer/schedulers/base.py:82

↓ 10 callersMethod_make_layer

(self, block, inplanes, planes, blocks, stride=1, expansion=None)

src/netspresso_trainer/models/full/experimental/pidnet.py:123

↓ 10 callersFunctionnatural_key

See http://www.codinghorror.com/blog/archives/001018.html

src/netspresso_trainer/dataloaders/utils/misc.py:43

↓ 10 callersFunctiontransform_bbox

(bboxes: Union[Tensor, Proxy], indicator="xywh -> xyxy", image_shape: Op

src/netspresso_trainer/utils/bbox_utils.py:37

↓ 9 callersMethodresult

(self, phase='train')

src/netspresso_trainer/metrics/base.py:55

↓ 8 callersMethod__init__

(self, pool_size=3)

src/netspresso_trainer/models/op/base_metaformer.py:37

↓ 8 callersMethodcalc

(self, out: Dict, target: Union[torch.Tensor, Dict[str, torch.Tensor]], phase='train', **kwargs: Any)

src/netspresso_trainer/losses/builder.py:79

↓ 7 callersMethod__init__

( self, size: Union[int, List], fill: Union[int, List], )

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:214

↓ 7 callersMethodstate_dict

(self)

src/netspresso_trainer/schedulers/base.py:70

↓ 6 callersMethod__init__

( self, params, optimizer_conf, )

src/netspresso_trainer/optimizers/custom.py:101

↓ 6 callersMethod__init__

(self, hidden_size, num_attention_heads, attention_hidden_size, attention_dropout_prob, atten

src/netspresso_trainer/models/backbones/experimental/efficientformer.py:114

↓ 6 callersMethod__init__

(self, in_channels, out_channels, BatchNorm=nn.BatchNorm2d)

src/netspresso_trainer/models/op/pidnet.py:387

↓ 6 callersFunctionbuild_model

(model_conf, num_classes, devices, distributed)

src/netspresso_trainer/models/builder.py:96

↓ 6 callersFunctionget_aug_params

(value, center=0)

src/netspresso_trainer/dataloaders/augmentation/custom/mosaic.py:55

↓ 5 callersMethod__init__

(self, hidden_size, num_attention_heads, attention_dropout_prob, sequence_reduction_ratio, in

src/netspresso_trainer/models/backbones/experimental/mixtransformer.py:97

↓ 5 callersMethodbackward

(self, grad_scaler: torch.cuda.amp.GradScaler)

src/netspresso_trainer/losses/builder.py:66

↓ 5 callersMethodget_params

( img, scale: Tuple[float, float], ratio: Tuple[float, float], value: Optional[int] = None )

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:431

↓ 5 callersFunctionis_single_task_model

(model_conf: omegaconf.DictConfig)

src/netspresso_trainer/models/utils.py:259

↓ 5 callersFunctionparse_args_netspresso

(with_gpus=False, isTrain=True)

src/netspresso_trainer/utils/engine_utils.py:47

↓ 5 callersFunctionrope

Applies Rotary Position Embedding to input tensor. Args: x (torch.Tensor): Input tensor. dim (int | list[int]): The spatial dimen

src/netspresso_trainer/models/heads/pose_estimation/experimental/rtmcc.py:35

↓ 5 callersFunctionset_arguments

( data: Union[Path, str], augmentation: Union[Path, str], model: Union[Path, str], logging: Un

src/netspresso_trainer/utils/engine_utils.py:137

↓ 5 callersMethodtrain

(self)

src/netspresso_trainer/pipelines/train.py:157

↓ 5 callersMethodtrunc_normal_

Copied from torch.nn.init._no_grad_trunc_normal_ This is instead of scipy.stats.tuncnorm (Not to add dependency)

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:857

↓ 4 callersMethod__init__

(self, num_classes, classwise_analysis, **kwargs)

src/netspresso_trainer/metrics/detection/metric.py:306

↓ 4 callersMethod__init__

(self, conf_model, backbone, neck, head, freeze_backbone: bool = False)

src/netspresso_trainer/models/base.py:31

↓ 4 callersMethod__init__

( self, task: str, params: Optional[DictConfig] = None, stage_params: Optional

src/netspresso_trainer/models/backbones/experimental/mobilevit.py:383

↓ 4 callersMethod__init__

(self, input_dim, hidden_dim, output_dim, num_layers, act='relu')

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:94

↓ 4 callersFunction_as_grid

(image_batch, rows: int, cols: int)

src/netspresso_trainer/loggers/tensorboard.py:144

↓ 4 callersMethod_get_src_permutation_idx

(self, indices)

src/netspresso_trainer/losses/detection/rtdetr.py:253

↓ 4 callersMethod_log_metric

(self, key: str, value, mode: Optional[str] = None)

src/netspresso_trainer/loggers/mlflow.py:81

↓ 4 callersMethod_save_model

(self, model, epoch: int, model_name_tag: str, logging_dir: Path)

src/netspresso_trainer/pipelines/train.py:306

↓ 4 callersFunctionbox_area

(box)

src/netspresso_trainer/metrics/detection/metric.py:49

↓ 4 callersFunctionbuild_dataloader

(conf, task: str, model_name: str, dataset, phase, profile=False)

src/netspresso_trainer/dataloaders/builder.py:176

↓ 4 callersFunctionchange_tab_to

(destination=None)

demo/tab/home/main.py:19

↓ 4 callersFunctioncreate_loader

( dataset, dataset_name, logger, batch_size=1, is_training=False,

src/netspresso_trainer/dataloaders/utils/loader.py:83

↓ 4 callersMethoddevice

(self)

src/netspresso_trainer/models/base.py:59

↓ 4 callersMethodend_record

(self, name)

src/netspresso_trainer/utils/record.py:118

↓ 4 callersFunctionget_detection_label

(label_file: Path)

src/netspresso_trainer/dataloaders/utils/misc.py:48

↓ 4 callersFunctionget_model_format

(model_conf: omegaconf.DictConfig)

src/netspresso_trainer/models/utils.py:267

↓ 4 callersMethodlog_results

( self, prefix: Literal['training', 'validation', 'evaluation', 'inference'], epoch: O

src/netspresso_trainer/pipelines/base.py:53

↓ 4 callersMethodsave_summary

(self, end_training=False, status="", error_stats="")

src/netspresso_trainer/pipelines/train.py:377

↓ 4 callersFunctionseparate_no_weights_decay

(params: set, module_config: DictConfig)

src/netspresso_trainer/optimizers/builder.py:54

↓ 4 callersFunctionset_logger

(level: str = 'INFO', distributed=False)

src/netspresso_trainer/utils/logger.py:77

↓ 4 callersFunctionset_training_targets

(train_model, targets: dict)

src/netspresso_trainer/models/utils.py:285

↓ 4 callersMethodstart_record

(self, name)

src/netspresso_trainer/utils/record.py:106

↓ 3 callersMethod__init__

(self, hidden_size, num_attention_heads, attention_dropout_prob, intermediate_size, hidden_dropout_prob, layer

src/netspresso_trainer/models/backbones/experimental/vit.py:95

↓ 3 callersMethod__init__

(self, in_planes, out_planes, num_groups)

src/netspresso_trainer/models/backbones/experimental/mixnet.py:45

↓ 3 callersMethod__init__

(self, dim, init_value=1., trainable=True)

src/netspresso_trainer/models/heads/pose_estimation/experimental/rtmcc.py:92

↓ 3 callersMethod__init__

(self, anchor_grid, scaler, reg_max: int)

src/netspresso_trainer/losses/detection/yolov9.py:293

↓ 3 callersMethod__init__

(self, ignore_index=IGNORE_INDEX_NONE_VALUE, weight=None, **kwargs)

src/netspresso_trainer/losses/segmentation/pidnet.py:154

↓ 3 callersFunction_as_image_array

(img: np.ndarray)

src/netspresso_trainer/loggers/visualizer.py:184

↓ 3 callersMethod_fuse_bn_tensor

(self, branch: Union[ConvLayer, nn.BatchNorm2d, nn.Identity])

src/netspresso_trainer/models/op/custom.py:255

↓ 3 callersMethod_get_transformed

(self, image, label, mask, bbox, keypoint, visualize_for_debug, dataset)

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:50

↓ 3 callersMethod_make_single_layer

(self, block, inplanes, planes, stride=1, expansion=None)

src/netspresso_trainer/models/full/experimental/pidnet.py:143

↓ 3 callersMethod_render

(self, df: Optional[pd.DataFrame] = None)

demo/func/experiments.py:121

↓ 3 callersMethod_set_aux_loss

(self, outputs_class, outputs_coord)

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:650

↓ 3 callersFunctionadd_file_handler

(log_filepath: str, distributed: bool)

src/netspresso_trainer/utils/logger.py:62

↓ 3 callersFunctionaverage_precisions_per_class

Compute the average precision, given the recall and precision curves. Source: https://github.com/rafaelpadilla/Object-Detection-Metrics.

src/netspresso_trainer/metrics/detection/metric.py:130

↓ 3 callersFunctionbitget

(byteval, idx)

tools/device_runtime/visualizers/detection_visualizer.py:22

↓ 3 callersFunctionbitget

(byteval, idx)

src/netspresso_trainer/loggers/visualizer.py:27

↓ 3 callersFunctionbox_iou

(boxes1, boxes2)

src/netspresso_trainer/losses/detection/rtdetr.py:32

↓ 3 callersFunctionbuild_dataset

( conf_data: DictConfig, conf_augmentation: DictConfig, task: str, model_name: str, distri

src/netspresso_trainer/dataloaders/builder.py:88

↓ 3 callersFunctionbuild_logger

( conf, task: str, model_name: str, step_per_epoch: int, class_map: Dict[int, str], re

src/netspresso_trainer/loggers/builder.py:23

↓ 3 callersFunctionbuild_optimizer

( model, single_task_model: bool, optimizer_conf: DictConfig, )

src/netspresso_trainer/optimizers/builder.py:114

↓ 3 callersFunctionbuild_pipeline

( pipeline_type: str, conf: DictConfig, task: str, model_name: str, model: nn.Module,

src/netspresso_trainer/pipelines/builder.py:63

↓ 3 callersFunctionbuild_scheduler

(optimizer, training_conf)

src/netspresso_trainer/schedulers/builder.py:22

↓ 3 callersMethoddropdown_selected

(dropdown_input)

demo/func/experiments.py:145

↓ 3 callersFunctionget_device

(x: Union[torch.Tensor, nn.Module])

src/netspresso_trainer/utils/environment.py:61

↓ 3 callersFunctionget_gpus_from_parser_and_config

( gpus: Optional[Union[List, int]], conf_environment: DictConfig )

src/netspresso_trainer/utils/engine_utils.py:171

↓ 3 callersMethodget_loss

(self, loss, outputs, targets, indices, num_boxes, **kwargs)

src/netspresso_trainer/losses/detection/rtdetr.py:265

↓ 3 callersMethodget_metric_with_all_outputs

(self, outputs, phase: Literal['train', 'valid'], metric_factory)

src/netspresso_trainer/pipelines/task_processors/base.py:62

↓ 3 callersFunctionget_new_logging_dir

(output_root_dir, project_id, mode: Literal['training', 'evaluation', 'inference'])

src/netspresso_trainer/utils/engine_utils.py:183

↓ 3 callersFunctionget_params_and_flops

(model, sample_input: torch.Tensor)

src/netspresso_trainer/utils/stats.py:27

↓ 3 callersMethodget_reassigned_t_i

adjust T_i to finish at last epoch in overall schedule

src/netspresso_trainer/schedulers/cosine_warm_restart.py:103

↓ 3 callersFunctionload_checkpoint

(f: Union[str, Path])

src/netspresso_trainer/utils/checkpoint.py:35

↓ 3 callersMethodload_data

(self)

src/netspresso_trainer/dataloaders/base.py:140

↓ 3 callersMethodload_state_dict

(self, state_dict: Dict[str, Any])

src/netspresso_trainer/schedulers/base.py:73

↓ 3 callersMethodlog_scalar

(self, key, value, mode='train')

src/netspresso_trainer/loggers/tensorboard.py:70

↓ 3 callersMethodpull_item

(self, index)

src/netspresso_trainer/dataloaders/detection.py:206

↓ 3 callersMethodsave_best

(self)

src/netspresso_trainer/pipelines/train.py:326

↓ 3 callersMethodsave_checkpoint

(self, epoch: int)

src/netspresso_trainer/pipelines/train.py:284

↓ 3 callersMethodsave_ndarray_as_image

(self, image_array: np.ndarray, filename: Union[str, Path], dataformats: Literal['HWC', 'CHW'] = 'HWC')

src/netspresso_trainer/loggers/image.py:32

↓ 3 callersFunctionset_device

(seed)

src/netspresso_trainer/utils/environment.py:31

↓ 3 callersMethodtranspose_for_scores

(self, x: Tensor, attention_head_size: int)

src/netspresso_trainer/models/op/base_metaformer.py:142

↓ 3 callersMethodtranspose_for_scores

(self, x: Tensor, num_head: int)

src/netspresso_trainer/models/op/base_metaformer.py:302

↓ 3 callersFunctionunpickle

(file)

tools/open_dataset_tool/cifar100.py:36

↓ 3 callersMethodupdate_groups

(self, values)

src/netspresso_trainer/schedulers/base.py:94

↓ 2 callersMethod__init__

( self, conf_data, conf_augmentation, model_name,

src/netspresso_trainer/dataloaders/classification.py:182

↓ 2 callersMethod__init__

( self, conf_data, conf_augmentation, model_name,

src/netspresso_trainer/dataloaders/segmentation.py:230

↓ 2 callersMethod__init__

( self, intermediate_features_dim: List[int], params: DictConfig, )

src/netspresso_trainer/models/necks/experimental/rtdetr_hybrid_encoder.py:103

↓ 2 callersMethod__init__

( self, num_classes: int, intermediate_features_dim: List[int], params: DictCo

src/netspresso_trainer/models/heads/detection/experimental/anchor_decoupled_head.py:34

↓ 2 callersMethod__init__

( self, num_classes: int, intermediate_features_dim: List[int], params: Dict

src/netspresso_trainer/models/heads/detection/experimental/yolo_fastest_head.py:29

↓ 2 callersMethod__init__

(self, **kwargs)

src/netspresso_trainer/losses/detection/retinanet.py:30

next →1–100 of 1,065, ranked by callers