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

↓ 2 callersMethod_add_noise

(self, lrs, t)

src/netspresso_trainer/schedulers/base.py:103

↓ 2 callersMethod_augmentation_space

(self, num_bins: int)

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:625

↓ 2 callersMethod_clear_epoch_start

(self)

src/netspresso_trainer/losses/builder.py:54

↓ 2 callersMethod_clear_step_start

(self)

src/netspresso_trainer/losses/builder.py:50

↓ 2 callersFunction_convert_graphmodule

(model: nn.Module)

src/netspresso_trainer/utils/fx.py:23

↓ 2 callersMethod_convert_scalar_as_readable

(self, scalar_dict: Dict)

src/netspresso_trainer/loggers/base.py:101

↓ 2 callersMethod_crop_bbox

(self, bbox, i, j, h, w)

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:321

↓ 2 callersMethod_crop_bbox

(self, bbox, i, j, h, w)

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:369

↓ 2 callersMethod_forward

GAU Forward function.

src/netspresso_trainer/models/heads/pose_estimation/experimental/rtmcc.py:257

↓ 2 callersMethod_forward

(self, out: torch.Tensor, target: torch.Tensor)

src/netspresso_trainer/losses/segmentation/pidnet.py:44

↓ 2 callersMethod_generate_anchors

(self, spatial_shapes=None, grid_size=0.05,

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:509

↓ 2 callersMethod_get_3rd_point

To calculate the affine matrix, three pairs of points are required. This function is used to get the 3rd point, given 2D points a & b

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:893

↓ 2 callersMethod_log_image

(self, key: str, value: Union[np.ndarray, torch.Tensor], mode)

src/netspresso_trainer/loggers/tensorboard.py:63

↓ 2 callersMethod_log_scalar

(self, key: str, value, mode)

src/netspresso_trainer/loggers/tensorboard.py:58

↓ 2 callersMethod_make_layer

(self, block: Type[Union[BasicBlock, Bottleneck]], planes: int, blocks: int, stride: int =

src/netspresso_trainer/models/backbones/core/resnet.py:137

↓ 2 callersMethod_round_filters

(self, filters, width_multi)

src/netspresso_trainer/models/backbones/experimental/mixnet.py:269

↓ 2 callersFunction_validate_file

(file_path: Path)

src/netspresso_trainer/utils/checkpoint.py:26

↓ 2 callersFunction_voc_color_map

(N=256, normalized=False)

src/netspresso_trainer/loggers/visualizer.py:26

↓ 2 callersFunctionauto_pad

Auto Padding for the convolution blocks

src/netspresso_trainer/models/op/custom.py:56

↓ 2 callersFunctionbox_xyxy_to_cxcywh

(x)

src/netspresso_trainer/models/op/denoising.py:35

↓ 2 callersMethodbuild_2d_sincos_position_embedding

src/netspresso_trainer/models/necks/experimental/rtdetr_hybrid_encoder.py:187

↓ 2 callersFunctionbuild_losses

(conf_model, **kwargs)

src/netspresso_trainer/losses/builder.py:97

↓ 2 callersFunctionbuild_metrics

(task: str, model_conf, metrics_conf, num_classes, **kwargs)

src/netspresso_trainer/metrics/builder.py:23

↓ 2 callersFunctioncalculate_iou

(bbox1, bbox2, metrics="iou")

src/netspresso_trainer/losses/detection/yolov9.py:36

↓ 2 callersFunctioncxcywh2cxcywhn

(cx, cy, w, h, img_w, img_h)

tools/open_dataset_tool/coco2017.py:47

↓ 2 callersMethodend

(self)

src/netspresso_trainer/utils/record.py:87

↓ 2 callersFunctionevaluation_common

( conf: DictConfig, task: str, model_name: str, logging_dir: Path, log_level: Literal['DEB

src/netspresso_trainer/evaluator_common.py:31

↓ 2 callersFunctiongeneralized_box_iou

Generalized IoU from https://giou.stanford.edu/ The boxes should be in [x0, y0, x1, y1] format Returns a [N, M] pairwise matrix, where

src/netspresso_trainer/losses/detection/rtdetr.py:48

↓ 2 callersFunctiongenerate_anchors

This implementation is based on https://github.com/WongKinYiu/YOLO/blob/main/yolo/utils/bounding_box_utils.py.

src/netspresso_trainer/utils/bbox_utils.py:11

↓ 2 callersFunctiongenerate_edge

(label: np.ndarray)

src/netspresso_trainer/dataloaders/augmentation/transforms.py:39

↓ 2 callersMethodget_assignments

( self, batch_idx, num_gt, gt_bboxes_per_image, gt_classes, bb

src/netspresso_trainer/losses/detection/yolox.py:284

↓ 2 callersMethodget_best_epoch

(self)

src/netspresso_trainer/pipelines/train.py:134

↓ 2 callersFunctionget_format

(level: str, distributed: bool = False)

src/netspresso_trainer/utils/logger.py:46

↓ 2 callersMethodget_losses

( self, imgs, x_shifts, y_shifts, expanded_strides, target,

src/netspresso_trainer/losses/detection/yolox.py:116

↓ 2 callersMethodget_predictions

(self, results, class_map)

src/netspresso_trainer/pipelines/task_processors/base.py:66

↓ 2 callersFunctioninverse_sigmoid

(x: torch.Tensor, eps: float=1e-5)

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:81

↓ 2 callersFunctionis_normalized

(fmt: str)

src/netspresso_trainer/utils/bbox_utils.py:40

↓ 2 callersFunctionload_from_checkpoint

( model: nn.Module, model_checkpoint: Optional[Union[str, Path]], load_checkpoint_head: bool, )

src/netspresso_trainer/models/utils.py:223

↓ 2 callersMethodlog_hparams

(self, hp_omegaconf: OmegaConf)

src/netspresso_trainer/loggers/mlflow.py:191

↓ 2 callersFunctionmake_divisible

This function is taken from the original tf repo. It ensures that all layers have a channel number that is divisible by 8 It can be seen

src/netspresso_trainer/models/op/custom.py:33

↓ 2 callersFunctionparse_gpu_ids

Parse comma-separated GPU IDs and return as a list of integers.

src/netspresso_trainer/utils/engine_utils.py:120

↓ 2 callersMethodrel_pos_bias

Add relative position bias.

src/netspresso_trainer/models/heads/pose_estimation/experimental/rtmcc.py:243

↓ 2 callersMethodreset

(self)

src/netspresso_trainer/metrics/segmentation/metric.py:36

↓ 2 callersMethodreset_values

(self)

src/netspresso_trainer/metrics/base.py:44

↓ 2 callersFunctionsave_exir

(model: nn.Module, f: Union[str, Path], sample_input: Tensor)

src/netspresso_trainer/utils/exir.py:29

↓ 2 callersFunctionsave_onnx

(model: nn.Module, f: Union[str, Path], sample_input: Tensor, opset_version)

src/netspresso_trainer/utils/onnx.py:43

↓ 2 callersMethodset_training_targets

Set internal training targets before training step

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:663

↓ 2 callersMethodslider_selected

(slider_input)

demo/func/experiments.py:149

↓ 2 callersFunctiontrain_common

( conf: DictConfig, task: str, model_name: str, is_graphmodule_training: bool, logging_dir

src/netspresso_trainer/trainer_common.py:31

↓ 2 callersMethodtrain_step

(self, train_model, batch)

src/netspresso_trainer/pipelines/task_processors/base.py:50

↓ 2 callersFunctiontrain_with_yaml_impl

(gpus: Optional[Union[List, int]], data: Union[Path, str], augmentation: Union[Path, str],

src/netspresso_trainer/trainer_main.py:71

↓ 2 callersFunctiontransforms_check

(transforms)

src/netspresso_trainer/dataloaders/augmentation/transforms.py:49

↓ 2 callersFunctiontxtywh2cxcywh

(top_left_x, top_left_y, width, height)

tools/open_dataset_tool/coco2017.py:39

↓ 2 callersMethodvalid_step

(self, eval_model, batch)

src/netspresso_trainer/pipelines/task_processors/base.py:54

↓ 2 callersMethodwith_pos_embed

(self, tensor, pos)

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:250

↓ 2 callersMethodxywhn2xyxy

(original: np.ndarray, w: int, h: int, padw=0, padh=0)

src/netspresso_trainer/dataloaders/detection.py:111

↓ 1 callersMethod__call__

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

src/netspresso_trainer/losses/builder.py:82

↓ 1 callersMethod__init__

(self, num_classes, classwise_analysis, **kwargs)

src/netspresso_trainer/metrics/classification/metric.py:50

↓ 1 callersMethod__init__

(self, num_classes, classwise_analysis, ignore_index=IGNORE_INDEX_NONE_VALUE, **kwargs)

src/netspresso_trainer/metrics/segmentation/metric.py:65

↓ 1 callersMethod__init__

(self, conf_data, train_valid_split_ratio)

src/netspresso_trainer/dataloaders/detection.py:39

↓ 1 callersMethod__init__

(self, conf_data, conf_augmentation, model_name, root, split, transform)

src/netspresso_trainer/dataloaders/base.py:82

↓ 1 callersMethod__init__

(self, conf_data, train_valid_split_ratio)

src/netspresso_trainer/dataloaders/pose_estimation.py:36

↓ 1 callersMethod__init__

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

src/netspresso_trainer/models/necks/experimental/fpn.py:33

↓ 1 callersMethod__init__

(self, in_channel, hidden_channel, out_channel, kernel_size, stride, norm_type, act_type)

src/netspresso_trainer/models/backbones/experimental/mobilenetv4.py:37

↓ 1 callersMethod__init__

(self, in_channels: int, hidden_channels: int, num_classes:

src/netspresso_trainer/models/heads/detection/experimental/yolo_head.py:43

↓ 1 callersMethod__init__

(self, feature_dim: int, num_classes: int, params: DictConfig)

src/netspresso_trainer/models/heads/classification/core/fc.py:31

↓ 1 callersMethod__init__

(self, weight: Optional[Tensor]=None, size_average=None, ignore_index: int=-100, reduce=None,

src/netspresso_trainer/losses/common.py:26

↓ 1 callersMethod__init__

(self, reduction="none", loss_type="iou")

src/netspresso_trainer/losses/detection/yolox.py:466

↓ 1 callersMethod__init__

Create the criterion. Parameters: num_classes: number of object categories, omitting the special no-object category m

src/netspresso_trainer/losses/detection/rtdetr.py:138

↓ 1 callersMethod__len__

(self)

src/netspresso_trainer/dataloaders/base.py:110

↓ 1 callersMethod_apply_cutmix_transform

(self, image_tensor, target_tensor)

src/netspresso_trainer/dataloaders/augmentation/custom/mixing.py:183

↓ 1 callersMethod_apply_mixup_transform

(self, image_tensor, target_tensor)

src/netspresso_trainer/dataloaders/augmentation/custom/mixing.py:100

↓ 1 callersMethod_as_numpy

(self, value: Union[np.ndarray, torch.Tensor, list])

src/netspresso_trainer/loggers/tensorboard.py:40

↓ 1 callersMethod_assert_argument

(self, kwargs)

src/netspresso_trainer/losses/builder.py:63

↓ 1 callersMethod_build_heads

( self, intermediate_features_dim: List[int], act_type: str, r

src/netspresso_trainer/models/heads/detection/experimental/yolo_head.py:128

↓ 1 callersMethod_build_input_proj_layer

(self, feat_channels)

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:459

↓ 1 callersMethod_build_losses

(self, **kwargs)

src/netspresso_trainer/losses/builder.py:39

↓ 1 callersMethod_build_network

(self)

src/netspresso_trainer/models/backbones/experimental/shufflenetv2.py:84

↓ 1 callersMethod_calc_distances

(self, preds, gts, mask, norm_factor)

src/netspresso_trainer/metrics/pose_estimation/metric.py:46

↓ 1 callersMethod_calculate_noise

(self, t)

src/netspresso_trainer/schedulers/base.py:119

↓ 1 callersMethod_convert

(self, gray_image)

src/netspresso_trainer/loggers/visualizer.py:139

↓ 1 callersMethod_convert_images_as_readable

(self, samples: Union[Dict, List])

src/netspresso_trainer/loggers/base.py:120

↓ 1 callersMethod_create_main_branch

(self)

src/netspresso_trainer/models/op/custom.py:1002

↓ 1 callersMethod_create_proj_branch

(self)

src/netspresso_trainer/models/op/custom.py:1010

↓ 1 callersMethod_crop_bbox

(self, bbox, i, j, h, w)

src/netspresso_trainer/dataloaders/augmentation/custom/image_proc.py:546

↓ 1 callersFunction_custom_logger

(level: str, distributed: bool)

src/netspresso_trainer/utils/logger.py:69

↓ 1 callersMethod_distance_acc

(self, distances, thr)

src/netspresso_trainer/metrics/pose_estimation/metric.py:58

↓ 1 callersMethod_end_record

(self, name)

src/netspresso_trainer/utils/record.py:112

↓ 1 callersMethod_freeze_backbone

(self)

src/netspresso_trainer/models/base.py:50

↓ 1 callersFunction_get_dataframe_from_experiment_list

(experiment_list: List[ExperimentSummary])

demo/func/experiments.py:101

↓ 1 callersMethod_get_decoder_input

(self, memory, spatial_shapes,

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:539

↓ 1 callersMethod_get_encoder_input

(self, feats)

src/netspresso_trainer/models/heads/detection/experimental/rtdetr_head.py:480

↓ 1 callersFunction_get_experiment_list

(experiment_dir)

demo/func/experiments.py:54

↓ 1 callersFunction_get_lr_list

( optimizer: torch.optim.Optimizer, scheduler: torch.optim.lr_scheduler._LRScheduler, total_epochs: in

demo/func/scheduler.py:12

↓ 1 callersMethod_get_rasterized_hparam

(self, hparams)

src/netspresso_trainer/loggers/tensorboard.py:84

↓ 1 callersMethod_get_stage_out_channels

(self, model_size: str)

src/netspresso_trainer/models/backbones/experimental/shufflenetv2.py:63

↓ 1 callersMethod_get_valid_records

(self, best_model_criterion)

src/netspresso_trainer/pipelines/train.py:117

↓ 1 callersMethod_global_pool

(self, x: Tensor, dims: List)

src/netspresso_trainer/models/op/custom.py:768

↓ 1 callersMethod_import_tflite

(self)

src/netspresso_trainer/models/base.py:200

↓ 1 callersMethod_initialize_weights

(self)

src/netspresso_trainer/models/backbones/experimental/mixnet.py:245

← previous next →101–200 of 1,065, ranked by callers