MCPcopy Create free account

hub / github.com/NVlabs/prismer / functions

Functions669 in github.com/NVlabs/prismer

⨍ Functions 669◇ Types & classes 166
↓ 78 callersMethodload
Load model from file. Args: path (str): file path
experts/depth/base_model.py:5
↓ 48 callersMethodto
(self, device)
experts/segmentation/mask2former/utils/misc.py:30
↓ 39 callersMethodget
(self)
experts/obj_detection/unidet/predictor.py:196
↓ 11 callersMethoddevice
(self)
experts/segmentation/mask2former/maskformer_model.py:164
↓ 11 callersFunctionget_expert_labels
(data_path, label_path, image_path, dataset, experts)
dataset/utils.py:74
↓ 10 callersMethod__init__
(self, config)
model/modules/roberta.py:187
↓ 10 callersFunctionget_activation
(name)
experts/depth/vit.py:12
↓ 9 callersFunction_conv3x3_bn_relu
(in_channels, out_channels, dilation=1)
experts/ocr_detection/charnet/modeling/model.py:21
↓ 9 callersMethoddecode
(char_scores)
experts/ocr_detection/charnet/modeling/postprocessing.py:233
↓ 9 callersFunctionget_label_path
(file_name, expert_name, with_suffix=False)
demo_vis.py:20
↓ 8 callersFunctioncreate_loader
(dataset, batch_size, num_workers, train, collate_fn=None)
dataset/__init__.py:36
↓ 8 callersFunctionget_attention
(name)
experts/depth/vit.py:22
↓ 8 callersFunctionnorm_normalize
(norm_out)
experts/normal/models/submodules/submodules.py:64
↓ 7 callersMethod__init__
( self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.0 )
experts/segmentation/mask2former/modeling/backbone/swin.py:24
↓ 6 callersMethod__init__
(self)
experts/edge/model.py:164
↓ 6 callersMethod__init__
Args: stem (nn.Module): a stem module stages (list[list[ResNetBlock]]): several (typically 4) stages,
experts/obj_detection/unidet/modeling/backbone/resnest.py:534
↓ 6 callersFunctionaffine_transform
(pair, affine_params)
dataset/randaugment.py:19
↓ 6 callersFunctionload_expert_model
(task=None)
experts/model_bank.py:11
↓ 6 callersMethodwith_pos_embed
(self, tensor, pos: Optional[Tensor])
experts/segmentation/mask2former/modeling/transformer_decoder/transformer.py:258
↓ 5 callersMethod__init__
( self, stem_module, stage_specs, transformation_module,
experts/ocr_detection/charnet/modeling/backbone/resnet.py:80
↓ 5 callersFunctioncreate_dataset
(dataset, config)
dataset/__init__.py:15
↓ 5 callersFunctioninterpolate_pos_embed
(orig_pos_embed, target_len)
model/modules/utils.py:34
↓ 4 callersMethod__init__
Init. Args: scale_factor (float): scaling mode (str): interpolation mode
experts/depth/blocks.py:141
↓ 4 callersMethod__init__
( self, d_model=512, nhead=8, num_encoder_layers=6, num_decoder_layers
experts/segmentation/mask2former/modeling/transformer_decoder/transformer.py:20
↓ 4 callersMethod__init__
(self, input_dim, hidden_dim, output_dim, num_layers)
experts/segmentation/mask2former/modeling/transformer_decoder/mask2former_transformer_decoder.py:195
↓ 4 callersFunction_make_fusion_block
(features, use_bn)
experts/depth/models.py:15
↓ 4 callersFunction_make_scratch
(in_shape, out_shape, groups=1, expand=False)
experts/depth/blocks.py:67
↓ 4 callersFunction_make_vit_b16_backbone
( model, features=[96, 192, 384, 768], size=[384, 384], hooks=[2, 5, 8, 11], vit_features=
experts/depth/vit.py:221
↓ 4 callersMethodbackward
(ctx, grad)
experts/ocr_detection/charnet/modeling/layers/misc.py:24
↓ 4 callersFunctioncosine_lr_schedule
Decay the learning rate
utils.py:13
↓ 4 callersMethodforward_features
(self, features)
experts/segmentation/mask2former/modeling/pixel_decoder/fpn.py:136
↓ 4 callersMethodlosses
(self, predictions, targets)
experts/segmentation/mask2former/modeling/meta_arch/per_pixel_baseline.py:114
↓ 4 callersFunctionpost_label_process
(inputs, labels_info)
dataset/utils.py:117
↓ 4 callersMethodput
(self, image)
experts/obj_detection/unidet/predictor.py:192
↓ 4 callersMethodtrain
Convert the model into training mode while keep layers freezed.
experts/segmentation/mask2former/modeling/backbone/swin.py:680
↓ 4 callersMethodwith_pos_embed
(self, tensor, pos: Optional[Tensor])
experts/segmentation/mask2former/modeling/transformer_decoder/mask2former_transformer_decoder.py:95
↓ 3 callersMethod__init__
(self, backbone=hourglass88())
experts/ocr_detection/charnet/modeling/model.py:113
↓ 3 callersMethod__init__
(self, start_index=1)
experts/depth/vit.py:58
↓ 3 callersMethod__init__
NOTE: this interface is experimental. Args: input_shape: shapes (channels and stride) of the input features t
experts/segmentation/mask2former/modeling/pixel_decoder/msdeformattn.py:167
↓ 3 callersMethod__init__
(self, *args, **kwargs)
experts/obj_detection/unidet/modeling/backbone/splat.py:23
↓ 3 callersFunction_get_activation_fn
Return an activation function given a string
experts/segmentation/mask2former/modeling/transformer_decoder/transformer.py:361
↓ 3 callersFunction_get_activation_fn
Return an activation function given a string
experts/segmentation/mask2former/modeling/transformer_decoder/mask2former_transformer_decoder.py:181
↓ 3 callersFunction_get_clones
(module, N)
experts/segmentation/mask2former/modeling/transformer_decoder/transformer.py:357
↓ 3 callersFunction_make_layer
(in_channels, out_channels, num_blocks, **kwargs)
experts/ocr_detection/charnet/modeling/backbone/hourglass.py:16
↓ 3 callersFunctionbuild_resnest_backbone
Create a ResNet instance from config. Returns: ResNet: a :class:`ResNet` instance.
experts/obj_detection/unidet/modeling/backbone/resnest.py:612
↓ 3 callersMethodfreeze
(self)
experts/obj_detection/unidet/modeling/backbone/resnest.py:50
↓ 3 callersMethodget_results
(self)
experts/obj_detection/unidet/evaluation/oideval.py:537
↓ 3 callersFunctionmap_back_unified_id
(results, map_back, reverse_id_mapping=None)
experts/obj_detection/unidet/evaluation/multi_dataset_evaluator.py:56
↓ 3 callersFunctionms_deform_attn_core_pytorch
(value, value_spatial_shapes, sampling_locations, attention_weights)
experts/segmentation/mask2former/modeling/pixel_decoder/ops/functions/ms_deform_attn_func.py:52
↓ 3 callersMethodprint_results
(self)
experts/obj_detection/unidet/evaluation/oideval.py:512
↓ 3 callersMethodrun
Wrapper function which calculates the results.
experts/obj_detection/unidet/evaluation/oideval.py:506
↓ 3 callersFunctionsample_points
(init_normal, gt_norm_mask, sampling_ratio, beta)
experts/normal/models/submodules/submodules.py:75
↓ 3 callersFunctionsetup_cfg
(args)
experts/segmentation/utils.py:6
↓ 3 callersFunctiontile
(x, dim, n_tile)
model/prismer_caption.py:115
↓ 3 callersFunctiontile
(x, dim, n_tile)
model/prismer_vqa.py:116
↓ 2 callersMethod__init__
(self, width: int, layers: int, heads: int)
model/modules/vit.py:63
↓ 2 callersMethod__init__
(self, in_channels, out_channels, stride=1)
experts/ocr_detection/charnet/modeling/backbone/hourglass.py:33
↓ 2 callersMethod__init__
NOTE: this interface is experimental. Args: input_shape: shapes (channels and stride) of the input features c
experts/segmentation/mask2former/modeling/pixel_decoder/fpn.py:40
↓ 2 callersMethod__init__
(self, args=None)
experts/normal/models/baseline.py:10
↓ 2 callersMethod__init__
(self, in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bia
experts/normal/models/submodules/submodules.py:47
↓ 2 callersMethod__init__
( self, unified_label_file, dataset_name, cfg, distributed, output_dir=None)
experts/obj_detection/unidet/evaluation/multi_dataset_evaluator.py:345
↓ 2 callersMethod_conv1x1_relu
(self, in_channels, out_channels)
experts/ocr_detection/charnet/modeling/backbone/decoder.py:27
↓ 2 callersFunction_evaluate_predictions_on_oid
(oid_gt, oid_results_path, eval_seg=False)
experts/obj_detection/unidet/evaluation/oideval.py:638
↓ 2 callersMethod_forward_box
(self, features, proposals, targets=None, dataset_source=-1)
experts/obj_detection/unidet/modeling/roi_heads/unified_roi_heads.py:66
↓ 2 callersMethod_forward_box
(self, features, proposals, targets=None, dataset_source=-1)
experts/obj_detection/unidet/modeling/roi_heads/split_roi_heads.py:85
↓ 2 callersMethod_freeze_stages
(self)
experts/segmentation/mask2former/modeling/backbone/swin.py:618
↓ 2 callersFunction_get_builtin_metadata
(cats)
experts/obj_detection/unidet/data/datasets/oid.py:810
↓ 2 callersMethod_get_gt_dt
Create gt, dt which are list of anns/dets. If use_cats is true only anns/dets corresponding to tuple (img_id, cat_id) will be used. El
experts/obj_detection/unidet/evaluation/oideval.py:241
↓ 2 callersMethod_get_src_permutation_idx
(self, indices)
experts/segmentation/mask2former/modeling/criterion.py:192
↓ 2 callersFunction_make_stage
( transformation_module, in_channels, bottleneck_channels, out_channels, block_count,
experts/ocr_detection/charnet/modeling/backbone/resnet.py:194
↓ 2 callersMethod_to_mask
(self, anns, lvis)
experts/obj_detection/unidet/evaluation/oideval.py:163
↓ 2 callersMethodattention
(self, x: torch.Tensor)
model/modules/vit.py:52
↓ 2 callersFunctionbuild_pixel_decoder
Build a pixel decoder from `cfg.MODEL.MASK_FORMER.PIXEL_DECODER_NAME`.
experts/segmentation/mask2former/modeling/pixel_decoder/fpn.py:21
↓ 2 callersFunctioncoco_caption_eval
(coco_gt_root, results_file)
utils.py:34
↓ 2 callersMethodevaluate
Run per image evaluation on given images and store results (a list of dict) in self.eval_imgs.
experts/obj_detection/unidet/evaluation/oideval.py:209
↓ 2 callersMethodforward
(self, x)
experts/ocr_detection/charnet/modeling/layers/misc.py:30
↓ 2 callersMethodforward_prediction_heads
(self, output, mask_features, attn_mask_target_size)
experts/segmentation/mask2former/modeling/transformer_decoder/mask2former_transformer_decoder.py:433
↓ 2 callersMethodget_loss
(self, loss, outputs, targets, indices, num_masks)
experts/segmentation/mask2former/modeling/criterion.py:204
↓ 2 callersFunctionget_readout_oper
(vit_features, features, use_readout, start_index=1)
experts/depth/vit.py:204
↓ 2 callersFunctionlink_val100
(dir_full, dir_100)
experts/obj_detection/datasets/prepare_panoptic_fpn.py:98
↓ 2 callersMethodlosses
enable advanced loss
experts/obj_detection/unidet/modeling/roi_heads/custom_fast_rcnn.py:160
↓ 2 callersFunctionmap_back_unified_id_novel_classes
(results, map_back, reverse_id_mapping=None)
experts/obj_detection/unidet/evaluation/multi_dataset_evaluator.py:67
↓ 2 callersFunctionnms
(boxes, overlapThresh, neighbourThresh=0.5, minScore=0, num_neig=0)
experts/ocr_detection/charnet/modeling/rotated_nms.py:13
↓ 2 callersFunctionpre_caption
(caption, max_words=50)
dataset/utils.py:163
↓ 2 callersFunctionpre_question
(question, max_words=50)
dataset/utils.py:177
↓ 2 callersMethodpredict_probs
(self, predictions, proposals)
experts/obj_detection/unidet/modeling/roi_heads/custom_fast_rcnn.py:104
↓ 2 callersFunctionregister_oid_instances
experts/obj_detection/unidet/data/datasets/register_oid.py:29
↓ 2 callersFunctionrotate_rect
(x1, y1, x2, y2, degree, center_x, center_y)
experts/ocr_detection/charnet/modeling/utils.py:11
↓ 2 callersMethodsigmoid_cross_entropy_loss
( self, pred_class_logits, gt_classes, use_advanced_loss=True)
experts/obj_detection/unidet/modeling/roi_heads/custom_fast_rcnn.py:115
↓ 2 callersFunctionwindow_partition
Args: x: (B, H, W, C) window_size (int): window size Returns: windows: (num_windows*B, window_size, window_size, C)
experts/segmentation/mask2former/modeling/backbone/swin.py:44
↓ 2 callersMethodwith_pos_embed
(self, tensor, pos: Optional[Tensor])
experts/segmentation/mask2former/modeling/transformer_decoder/transformer.py:179
↓ 2 callersMethodwith_pos_embed
(self, tensor, pos: Optional[Tensor])
experts/segmentation/mask2former/modeling/transformer_decoder/mask2former_transformer_decoder.py:37
↓ 1 callersFunctionCutoutAbs
(img, v)
dataset/randaugment.py:161
↓ 1 callersMethod__init__
(self, width: int, layers: int, heads: int, num_latents: int)
model/modules/resampler.py:40
↓ 1 callersMethod__init__
(self, word_bbox, word_bbox_score, text, text_score, char_scores)
experts/ocr_detection/charnet/modeling/postprocessing.py:39
↓ 1 callersMethod__init__
( self, head, features=256, backbone="vitb_rn50_384", readout="project
experts/depth/models.py:27
↓ 1 callersMethod__init__
(self, input_dim, hidden_dim, output_dim, num_layers)
experts/segmentation/mask2former/modeling/transformer_decoder/maskformer_transformer_decoder.py:177
↓ 1 callersMethod__init__
NOTE: this interface is experimental. Args: input_shape: shapes (channels and stride) of the input features n
experts/segmentation/mask2former/modeling/meta_arch/per_pixel_baseline.py:48
↓ 1 callersMethod__init__
Args: cfg (CfgNode): num_gpus (int): if 0, will run on CPU
experts/obj_detection/unidet/predictor.py:165
↓ 1 callersFunction_check_size_scale_factor
(dim)
experts/ocr_detection/charnet/modeling/layers/misc.py:83
next →1–100 of 669, ranked by callers