MCPcopy Create free account

hub / github.com/alipay/Ant-Multi-Modal-Framework / functions

Functions4,004 in github.com/alipay/Ant-Multi-Modal-Framework

⨍ Functions 4,004◇ Types & classes 941↳ Endpoints 1
↓ 2 callersMethodprocessPunctuation
(self, inText)
prj/Pink/pink/eval/vqa_tools/vqa_eval.py:249
↓ 2 callersMethodprocess_feature_embedding
( self, attr, sample_list, text_embedding_total, extra=[], batch_size_t=None )
antmmf/models/ant_mmf.py:207
↓ 2 callersMethodrandom_masking
(self, x, mask_ratio)
prj/M2_Encoder/vlmo/torchscale/architecture/encoder.py:271
↓ 2 callersMethodreduce_clips
(self, simi_logits, level)
prj/base_vtp/roi_univl/univl/model/univl_video_ret.py:345
↓ 2 callersMethodreduce_clips
(self, simi_logits, level)
prj/cnvid_vtp/roi_univl/univl/model/univl_video_ret.py:354
↓ 2 callersMethodregister_extractor
(self, feature_extractor)
antmmf/datasets/features/vision/feature_saver.py:173
↓ 2 callersFunctionreset
(nn)
antmmf/utils/init.py:51
↓ 2 callersMethodresidual_connection
(self, x, residual)
prj/M2_Encoder/vlmo/torchscale/architecture/encoder.py:110
↓ 2 callersFunctionresize
(image, target, size, max_size=None)
antmmf/datasets/processors/transforms/detection.py:101
↓ 2 callersMethodrestore
(self)
antmmf/models/nlp_adversarial.py:48
↓ 2 callersFunctionrotate_half
Rotates half the hidden dims of the input.
prj/M2_omni/models/modeling_llama_3d.py:191
↓ 2 callersFunctionsample_frames
(num_frames, total_frames, sample="random")
prj/M2_omni/models/m2omni_utils.py:175
↓ 2 callersFunctionsanitize
(type_repr: str)
antmmf/utils/inspector.py:26
↓ 2 callersMethodsave
(self, savepath)
antmmf/utils/visual_utils/visualization_utils.py:80
↓ 2 callersFunctionscatter_add
( src: torch.Tensor, index: torch.Tensor, dim: int = -1, out: Optional[torch.Tensor] = None,
antmmf/utils/scatter.py:39
↓ 2 callersFunctionscore_to_ranks
(scores)
antmmf/modules/metrics/recall_at_k.py:12
↓ 2 callersFunctionsetAccuracy
(accQA)
prj/Pink/pink/eval/eval_vqav2.py:149
↓ 2 callersMethodset_predictor_state
(self)
antmmf/predictors/base_predictor.py:37
↓ 2 callersFunctionshear_level_to_args
(MAX_LEVEL, replace_value)
prj/M2_Encoder/vlmo/transforms/randaugment.py:209
↓ 2 callersMethodsummarize
(self, *args, **kwargs)
antmmf/modules/metrics/f1.py:73
↓ 2 callersFunctiontensor2dtype
(tensors, dtype)
antmmf/utils/register_fp32.py:8
↓ 2 callersFunctiontiny_value_of_dtype
Returns a moderately tiny value for a given PyTorch data type that is used to avoid numerical issues such as division by zero. This is di
antmmf/modules/utils.py:40
↓ 2 callersMethodtrain
(self, mode: bool = True)
prj/Pink/pink/model/adapter.py:46
↓ 2 callersMethodtrain
(self)
antmmf/trainers/adv_trainer.py:70
↓ 2 callersFunctiontranslate_level_to_args
(translate_const, MAX_LEVEL, replace_value)
prj/M2_Encoder/vlmo/transforms/randaugment.py:219
↓ 2 callersFunctiontune_channel
(ch)
prj/M2_Encoder/vlmo/transforms/randaugment.py:16
↓ 2 callersMethodupdate_data
(self, data, prev_word_inds, next_word_inds, incomplete_inds)
antmmf/utils/text_utils.py:465
↓ 2 callersMethodupdate_global_step
(self, global_step)
antmmf/common/metrics_reporter.py:198
↓ 2 callersFunctionwindow_partition
(x, window_size)
antmmf/modules/vision/backbone/video_swin.py:95
↓ 2 callersFunctionwindow_partition
Args: x: (B, H, W, C) window_size (int): window size Returns: windows: (num_windows*B, window_size, window_size, C)
antmmf/modules/vision/backbone/cctt.py:130
↓ 2 callersMethodwith_default
Load default configuration from antmmf/common/defaults/configs/base.yaml if the configuration is not exist, and the configurations yo
antmmf/common/configuration.py:300
↓ 2 callersMethodwith_pos_embed
(self, tensor, pos: Optional[Tensor])
antmmf/modules/transformers/base.py:300
↓ 1 callersFunctionCutoutAbs
(img, v)
prj/M2_Encoder/vlmo/transforms/randaug.py:129
↓ 1 callersMethodVWM_get_cap_input
(self, caption_input, sample_list)
prj/snps3_vtp/roi_univl/univl/model/univl_video_ret.py:529
↓ 1 callersMethodVWM_model_similarity
(self, before_cap_embed, after_cap_embed, vis_input, output_dict=None, cal_cross=True)
prj/snps3_vtp/roi_univl/univl/model/univl_video_ret.py:599
↓ 1 callersMethod__check_input__
(self, edge_index, size)
antmmf/modules/message_passing/message_passing.py:82
↓ 1 callersMethod__collect__
(self, args, edge_index, size, kwargs)
antmmf/modules/message_passing/message_passing.py:120
↓ 1 callersMethod__del__
(self)
antmmf/common/metrics_reporter.py:93
↓ 1 callersMethod__getattr__
(self, key)
antmmf/common/configuration.py:586
↓ 1 callersMethod__implements__
(self, cls, func_name: str)
antmmf/utils/inspector.py:119
↓ 1 callersMethod__init__
(self, hidden_size)
prj/M2_Encoder/vlmo/modules/heads.py:5
↓ 1 callersMethod__init__
(self, seed)
prj/M2_Encoder/vlmo/torchscale/component/feedforward_network.py:15
↓ 1 callersMethod__init__
(self, module, dim=1)
prj/M2_Encoder/vlmo/torchscale/component/multiway_network.py:25
↓ 1 callersMethod__init__
( self, img_size=224, patch_size=16, in_chans=3, embed_dim=768,
prj/M2_Encoder/vlmo/torchscale/component/embedding.py:31
↓ 1 callersMethod__init__
( self, model_dim: int, num_experts: int, use_fp32=False, input_noise_
prj/M2_Encoder/vlmo/torchscale/component/xmoe/routing.py:149
↓ 1 callersMethod__init__
( self, args, embed_tokens=None, embed_positions=None, output_projection=None, is_encoder_decoder=Fals
prj/M2_Encoder/vlmo/torchscale/architecture/decoder.py:208
↓ 1 callersMethod__init__
( self, args, embed_tokens=None, embed_positions=None, output_projection=None, is_encoder_decoder=Fals
prj/M2_Encoder/vlmo/torchscale/architecture/encoder.py:172
↓ 1 callersMethod__init__
(self, config, detr_head=None, generation_head=None)
prj/snps3_vtp/roi_univl/univl/model/univl_pretrain.py:28
↓ 1 callersMethod__init__
(self, config: PinkConfig)
prj/Pink/pink/model/pink.py:127
↓ 1 callersMethod__init__
(self, config, detr_head=None, generation_head=None)
prj/dmae_vtp/roi_univl/univl/model/univl_pretrain.py:28
↓ 1 callersMethod__init__
(self, config, detr_head=None, generation_head=None)
prj/base_vtp/roi_univl/univl/model/univl_pretrain.py:28
↓ 1 callersMethod__init__
(self, config, detr_head=None, generation_head=None)
prj/cnvid_vtp/roi_univl/univl/model/univl_pretrain.py:28
↓ 1 callersMethod__init__
(self, parent=None, label_name=None, label_idx=-1)
antmmf/modules/utils.py:162
↓ 1 callersMethod__init__
(self, name: str = "multi_label_metric")
antmmf/modules/metrics/mce_accuracy.py:17
↓ 1 callersMethod__init__
(self, config: Configuration, **kwargs)
antmmf/modules/encoders/multimodal_encoder.py:18
↓ 1 callersMethod__init__
(self, encoder_type: str, *args, **kwargs)
antmmf/modules/encoders/image_feature_encoder.py:22
↓ 1 callersMethod__init__
(self, dim_in, dim_hid, dim_out, dropout, alpha, num_heads)
antmmf/modules/encoders/graph/gat_adj_matrix.py:69
↓ 1 callersMethod__init__
( self, fc_layer_num: int, in_dim: int, out_dim: int, hidden_dim: int
antmmf/modules/layers/feats_joint.py:36
↓ 1 callersMethod__init__
( self, in_dim: int, out_dim: int, hidden_dim: int = None, num_layers:
antmmf/modules/layers/mlp.py:36
↓ 1 callersMethod__init__
(self, in_dim, out_dim, hidden_dim)
antmmf/modules/layers/transform_layer.py:33
↓ 1 callersMethod__init__
(self, net, n_segment)
antmmf/modules/vision/temporal_shift.py:83
↓ 1 callersMethod__init__
(self, blocks_args=None, global_params=None)
antmmf/modules/vision/backbone/efficientnet.py:135
↓ 1 callersMethod__init__
(self, loss_list)
antmmf/modules/losses/losses.py:79
↓ 1 callersMethod__init__
(self, init_dict={})
antmmf/structures/sample.py:36
↓ 1 callersMethod__init__
(self, task_name)
antmmf/tasks/base_task.py:50
↓ 1 callersMethod__init__
:param annotation_path: :param anno_type: handle file with same suffix but using different analyzing method.
antmmf/datasets/database/annotated.py:39
↓ 1 callersMethod__init__
(self, row)
antmmf/datasets/database/video_database.py:23
↓ 1 callersMethod__init__
Feature Reader class for reading features. Note: Deprecation: ndim and image_feature will be deprecated later and the format will be
antmmf/datasets/features/feature_readers.py:25
↓ 1 callersMethod__init__
(self, s3d_model_path=None, *args, **kwargs)
antmmf/datasets/features/vision/video_feature.py:296
↓ 1 callersMethod__init__
(self, config, *args, **kwargs)
antmmf/datasets/processors/image_processors.py:145
↓ 1 callersMethod__init__
(self, config)
antmmf/models/bert.py:25
↓ 1 callersMethod__init__
(self, config, model)
antmmf/models/mm_adversarial.py:51
↓ 1 callersMethod__init__
(self, config)
antmmf/models/concat_bert.py:152
↓ 1 callersMethod__init__
(self, config)
antmmf/predictors/base_predictor.py:24
↓ 1 callersMethod__len__
(self)
prj/Pink/pink/eval/model_object365.py:77
↓ 1 callersMethod__len__
(self)
prj/Pink/pink/datasets/BaseDataset.py:75
↓ 1 callersMethod__len__
(self)
prj/Pink/pink/datasets/ProbMergeDataset.py:31
↓ 1 callersMethod__lift__
(self, src, edge_index, dim)
antmmf/modules/message_passing/message_passing.py:113
↓ 1 callersMethod_add_detectron_config
(self, cfg)
antmmf/datasets/features/vision/detectron_feature.py:295
↓ 1 callersMethod_add_extra_args_for_dataloader
(self, other_args={})
antmmf/common/test_reporter.py:195
↓ 1 callersMethod_add_overlap_chunk
(self, feats: np.ndarray, cache: dict = {})
prj/M2_omni/models/sanm_audio.py:464
↓ 1 callersMethod_adv_build_clean_sample
(self, sample_list)
antmmf/models/mm_adversarial.py:162
↓ 1 callersMethod_adv_grad
(self, noisy_samples, backward_param)
antmmf/models/mm_adversarial.py:432
↓ 1 callersMethod_adv_text_param
(self, noisy_samples)
antmmf/models/mm_adversarial.py:408
↓ 1 callersMethod_adv_update_noisy_sample
(self, noisy_samples, adv_info)
antmmf/models/mm_adversarial.py:467
↓ 1 callersMethod_adv_update_report
(self, report, noisy_sample)
antmmf/models/mm_adversarial.py:235
↓ 1 callersMethod_align_text_to_video_clips
(self, cap_embed, cap_mask, num_clip: int = 1)
prj/snps3_vtp/roi_univl/univl/model/univl_video_base.py:302
↓ 1 callersMethod_align_text_to_video_clips
(self, cap_embed, cap_mask, num_clip: int = 1)
prj/dmae_vtp/roi_univl/univl/model/univl_video_base.py:305
↓ 1 callersMethod_align_text_to_video_clips
(self, cap_embed, cap_mask, num_clip: int = 1)
prj/base_vtp/roi_univl/univl/model/univl_video_base.py:302
↓ 1 callersMethod_align_text_to_video_clips
(self, cap_embed, cap_mask, num_clip: int = 1)
prj/cnvid_vtp/roi_univl/univl/model/univl_video_base.py:302
↓ 1 callersMethod_attack_text
(self, noisy_samples, adv_info)
antmmf/models/mm_adversarial.py:444
↓ 1 callersMethod_backward
compute the inner product between the difference of parameter and a base parameter and a gradient, different from directly so
antmmf/models/mm_adversarial.py:332
↓ 1 callersMethod_build
This is used to build a dataset first time. Implement this method in your child dataset builder class. Args: dat
antmmf/datasets/base_dataset_builder.py:112
↓ 1 callersMethod_build_annotation_db
(self, database_cls=AnnotatedDatabase, **kwargs)
antmmf/datasets/base_dataset.py:171
↓ 1 callersMethod_build_detection_model
(self)
antmmf/scripts/features/extract_features_vmb.py:88
↓ 1 callersMethod_build_extractor
build model for feature extraction Returns: torch.nn.module object
antmmf/datasets/features/vision/base_extractor.py:43
↓ 1 callersMethod_build_features_db
(self, database_cls=FeaturesDatabase, **kwargs)
antmmf/datasets/base_dataset.py:207
↓ 1 callersMethod_build_image_db
(self, database_cls=ImageDatabase, **kwargs)
antmmf/datasets/base_dataset.py:181
↓ 1 callersMethod_build_input_for_multiple_choice
(self, context, choices)
prj/M2_Encoder/vlmo/tokenizer/tokenization_glm.py:73
↓ 1 callersMethod_build_postprocessor
some necessary procedures before getting final feature: do shape transform, format conversion .etc Returns:
antmmf/datasets/features/vision/base_extractor.py:52
← previousnext →601–700 of 4,004, ranked by callers