MCPcopy Create free account

hub / github.com/boheumd/MA-LMM / functions

Functions1,238 in github.com/boheumd/MA-LMM

⨍ Functions 1,238◇ Types & classes 310↳ Endpoints 9
↓ 1 callersMethod__getitem__
(self, index)
lavis/datasets/datasets/dialogue_datasets.py:71
↓ 1 callersMethod__getitem__
(self, index)
lavis/datasets/datasets/coco_vqa_datasets.py:37
↓ 1 callersMethod__getitem__
(self, index)
lavis/datasets/datasets/caption_datasets.py:44
↓ 1 callersMethod__getitem__
(self, index)
lavis/datasets/datasets/aok_vqa_datasets.py:38
↓ 1 callersMethod__getitem__
(self, index)
lavis/datasets/datasets/nlvr_datasets.py:65
↓ 1 callersMethod__init__
(self, vis_processor, text_processor, vis_root, ann_paths, num_frames, prompt='', split='train')
lavis/datasets/datasets/msrvtt_vqa_datasets.py:23
↓ 1 callersMethod__init__
vis_root (string): Root directory of images (e.g. coco/images/) ann_root (string): directory to store the annotation file spl
lavis/datasets/datasets/video_caption_datasets.py:15
↓ 1 callersMethod__init__
(self, vis_processor, text_processor, vis_root, ann_paths, num_frames, prompt='', split='train')
lavis/datasets/datasets/breakfast_cls_datasets.py:21
↓ 1 callersMethod__init__
vis_root (string): Root directory of images (e.g. coco/images/) ann_root (string): directory to store the annotation file spl
lavis/datasets/datasets/coco_caption_datasets.py:48
↓ 1 callersMethod__init__
(self, vis_processor, text_processor, vis_root, ann_paths, num_frames, prompt='', split='train')
lavis/datasets/datasets/activitynet_vqa_datasets.py:23
↓ 1 callersMethod__init__
(self, vis_processor, text_processor, vis_root, ann_paths, num_frames, prompt='', split='train')
lavis/datasets/datasets/msvd_vqa_datasets.py:23
↓ 1 callersMethod__init__
vis_root (string): Root directory of images (e.g. coco/images/) ann_root (string): directory to store the annotation file
lavis/datasets/datasets/base_dataset.py:16
↓ 1 callersMethod__init__
vis_root (string): Root directory of images (e.g. coco/images/) ann_root (string): directory to store the annotation file spl
lavis/datasets/datasets/avsd_dialogue_datasets.py:16
↓ 1 callersMethod__init__
(self, vis_processor, text_processor, vis_root, ann_paths, num_frames, prompt='', split='train')
lavis/datasets/datasets/msrvtt_caption_datasets.py:23
↓ 1 callersMethod__init__
(self, vis_processor, text_processor, vis_root, ann_paths, num_frames, prompt='', split='train')
lavis/datasets/datasets/coin_cls_datasets.py:21
↓ 1 callersMethod__init__
(self, vis_processor, text_processor, vis_root, ann_paths)
lavis/datasets/datasets/vqa_datasets.py:14
↓ 1 callersMethod__init__
vis_root (string): Root directory of images (e.g. coco/images/) ann_root (string): directory to store the annotation file
lavis/datasets/datasets/caption_datasets.py:29
↓ 1 callersMethod__init__
(self, vis_processor, text_processor, vis_root, ann_paths, num_frames, prompt='', split='train')
lavis/datasets/datasets/msvd_caption_datasets.py:23
↓ 1 callersMethod__init__
vis_root (string): Root directory of videos (e.g. LVU/videos/) ann_root (string): directory to store the gt_dict file
lavis/datasets/datasets/lvu_cls_datasets.py:21
↓ 1 callersMethod__init__
(self)
lavis/models/base_model.py:22
↓ 1 callersMethod__init__
(self, max_turns=3, use_caption=True)
lavis/processors/gpt_processors.py:47
↓ 1 callersMethod__init__
(self, image_size=224, mean=None, std=None)
lavis/processors/clip_processors.py:64
↓ 1 callersFunction_apply
(x)
lavis/datasets/data_utils.py:56
↓ 1 callersMethod_build_class_labels
(self)
lavis/datasets/datasets/snli_ve_datasets.py:37
↓ 1 callersMethod_build_opt_list
(self, opts)
lavis/common/config.py:52
↓ 1 callersMethod_convert_to_dot_list
(self, opts)
lavis/common/config.py:115
↓ 1 callersMethod_download_ann
Download annotation files if necessary. All the vision-language datasets should have annotations of unified format. storage_
lavis/datasets/builders/base_dataset_builder.py:93
↓ 1 callersMethod_download_data
(self)
lavis/datasets/builders/base_dataset_builder.py:89
↓ 1 callersMethod_download_vis
(self)
lavis/datasets/builders/base_dataset_builder.py:150
↓ 1 callersFunction_expand_mask
Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`.
lavis/models/blip2_models/modeling_llama.py:61
↓ 1 callersMethod_flip
(samples)
lavis/datasets/datasets/nlvr_datasets.py:44
↓ 1 callersMethod_generate_answers
(self, samples, num_beams=3, max_length=10, min_length=1)
lavis/models/blip_models/blip_vqa.py:237
↓ 1 callersMethod_get_answer_label
(self, answer)
lavis/datasets/datasets/video_vqa_datasets.py:38
↓ 1 callersFunction_get_google_drive_file_id
(url: str)
lavis/common/utils.py:141
↓ 1 callersMethod_lemmatize
(self, answers)
lavis/models/blip2_models/blip2_t5_instruct.py:698
↓ 1 callersMethod_lemmatize
(self, answers)
lavis/models/blip2_models/blip2_t5.py:314
↓ 1 callersMethod_lemmatize
(self, answers)
lavis/models/blip2_models/blip2_vicuna_instruct.py:759
↓ 1 callersMethod_load_checkpoint
Resume from a checkpoint.
lavis/runners/runner_iter.py:171
↓ 1 callersMethod_load_checkpoint
Resume from a checkpoint.
lavis/runners/runner_base.py:629
↓ 1 callersFunction_load_weights
Load weights from .npz checkpoints for official Google Brain Flax implementation
lavis/models/vit.py:289
↓ 1 callersFunction_make_causal_mask
Make causal mask used for bi-directional self-attention.
lavis/models/blip2_models/modeling_llama.py:43
↓ 1 callersFunction_natural_key
(string_)
lavis/models/clip_models/model.py:892
↓ 1 callersFunction_no_grad_trunc_normal_
(tensor, mean, std, a, b)
lavis/models/clip_models/timm_model.py:506
↓ 1 callersFunction_ntuple
(n)
lavis/models/clip_vit.py:190
↓ 1 callersMethod_prepare_decoder_attention_mask
(self, attention_mask, input_shape, inputs_embeds, past_key_values_length)
lavis/models/blip2_models/modeling_llama.py:460
↓ 1 callersMethod_progress
(self, cur_iters)
lavis/runners/runner_iter.py:68
↓ 1 callersMethod_rampup_factor
(self, epoch, iters, num_iters_per_epoch)
lavis/models/blip_models/blip_classification.py:74
↓ 1 callersMethod_rampup_factor
(self, epoch, iters, num_iters_per_epoch)
lavis/models/blip_models/blip_retrieval.py:113
↓ 1 callersMethod_rampup_factor
(self, epoch, iters, num_iters_per_epoch)
lavis/models/blip_models/blip_pretrain.py:111
↓ 1 callersMethod_rampup_factor
(self, epoch, iters, num_iters_per_epoch)
lavis/models/albef_models/albef_vqa.py:80
↓ 1 callersMethod_rampup_factor
(self, epoch, iters, num_iters_per_epoch)
lavis/models/albef_models/albef_classification.py:80
↓ 1 callersMethod_rampup_factor
(self, epoch, iters, num_iters_per_epoch)
lavis/models/albef_models/albef_nlvr.py:76
↓ 1 callersMethod_rampup_factor
(self, epoch, iters, num_iters_per_epoch)
lavis/models/albef_models/albef_pretrain.py:102
↓ 1 callersMethod_rampup_factor
(self, epoch, iters, num_iters_per_epoch)
lavis/models/albef_models/albef_retrieval.py:104
↓ 1 callersMethod_rank_answers
Generate the first token of answers using decoder and select ${num_ans_candidates} most probable ones. Then select answers from answe
lavis/models/blip_models/blip_vqa.py:277
↓ 1 callersMethod_relative_position_bucket
Adapted from Mesh Tensorflow: https://github.com/tensorflow/mesh/blob/0cb87fe07da627bf0b7e60475d59f95ed6b5be3d/mesh_tensorflow/transf
lavis/models/blip2_models/modeling_t5.py:393
↓ 1 callersMethod_reload_best_model
Load the best checkpoint for evaluation.
lavis/runners/runner_base.py:609
↓ 1 callersMethod_report_metrics
(scores_i2t, scores_t2i, txt2img, img2txt)
lavis/tasks/retrieval.py:54
↓ 1 callersMethod_report_metrics
(self, eval_result_file, split_name)
lavis/tasks/multimodal_classification.py:66
↓ 1 callersMethod_report_metrics
TODO: add other evaluation metrics for GQA
lavis/tasks/vqa_reading_comprehension.py:193
↓ 1 callersMethod_report_metrics_cls
(self, eval_result_file, split_name, dataset)
lavis/tasks/classification.py:112
↓ 1 callersMethod_report_metrics_coco
Use official VQA evaluation script to report metrics.
lavis/tasks/vqa.py:164
↓ 1 callersMethod_report_metrics_coco
(self, eval_result_file, split_name)
lavis/tasks/captioning.py:94
↓ 1 callersMethod_report_metrics_video_caption
(self, eval_result_file, split_name, dataset, dataset_name)
lavis/tasks/captioning.py:114
↓ 1 callersMethod_report_metrics_vqa
(self, result_file, split, dataset, dataset_name)
lavis/tasks/vqa.py:135
↓ 1 callersMethod_save_result_leaderboard
Saving the results in the format required for leaderboard evaluation.
lavis/tasks/vqa_reading_comprehension.py:232
↓ 1 callersMethodafter_evaluation
(self, val_result, split_name, epoch, dataset, **kwargs)
lavis/tasks/vqa.py:120
↓ 1 callersFunctionapply_memory_bank
(model, memory_bank_length, num_frames)
lavis/models/blip2_models/blip2.py:394
↓ 1 callersFunctionapply_rotary_pos_emb
(q, k, cos, sin, position_ids)
lavis/models/blip2_models/modeling_llama.py:134
↓ 1 callersFunctionapply_to_sample
(f, sample)
lavis/datasets/data_utils.py:52
↓ 1 callersFunctionbasic_clean
(text)
lavis/models/clip_models/tokenizer.py:70
↓ 1 callersMethodbefore_evaluation
(self, model, dataset, **kwargs)
lavis/tasks/base_task.py:74
↓ 1 callersMethodbpe
(self, token)
lavis/models/clip_models/tokenizer.py:111
↓ 1 callersMethodbuild
Create by split datasets inheriting torch.utils.data.Datasets. # build() can be dataset-specific. Overwrite to customize.
lavis/datasets/builders/video_qa_builder.py:133
↓ 1 callersMethodbuild
Create by split datasets inheriting torch.utils.data.Datasets. # build() can be dataset-specific. Overwrite to customize.
lavis/datasets/builders/base_dataset_builder.py:164
↓ 1 callersMethodbuild_attention_mask
(self)
lavis/models/clip_models/model.py:538
↓ 1 callersMethodbuild_dataset_config
(config)
lavis/common/config.py:89
↓ 1 callersMethodbuild_datasets
(self)
lavis/datasets/builders/base_dataset_builder.py:43
↓ 1 callersMethodbuild_model_config
(config, **kwargs)
lavis/common/config.py:57
↓ 1 callersFunctionbuild_rotary_pos_embed
NOTE: shape arg should include spatial dim only
lavis/models/clip_models/timm_model.py:449
↓ 1 callersMethodbuild_runner_config
(config)
lavis/common/config.py:85
↓ 1 callersFunctioncoco_caption_eval
(coco_gt_root, results_file, split)
lavis/tasks/captioning.py:156
↓ 1 callersFunctioncoco_dialogue_eval
(coco_gt_root, results_file, split)
lavis/tasks/dialogue.py:93
↓ 1 callersMethodcollater
(self, samples)
lavis/datasets/datasets/base_dataset.py:37
↓ 1 callersMethodcompute_bias
Compute binned relative position bias
lavis/models/blip2_models/modeling_t5.py:447
↓ 1 callersFunctioncompute_gradcam_batch
(model, visual_input, text_input, tokenized_text, block_num=6)
app/multimodal_search.py:191
↓ 1 callersFunctioncompute_sim_matrix
(model, data_loader, **kwargs)
lavis/models/blip2_models/blip2.py:420
↓ 1 callersMethodcompute_sim_matrix
(self, data_loader, **kwargs)
lavis/models/clip_models/model.py:720
↓ 1 callersFunctioncosine_lr_schedule
Decay the learning rate
lavis/common/optims.py:97
↓ 1 callersFunctioncreate_clip_vit_L
(img_size=224,use_checkpoint=False,precision="fp16")
lavis/models/clip_vit.py:233
↓ 1 callersFunctioncreate_eva_vit_g
(img_size=224,drop_path_rate=0.4,use_checkpoint=False,precision="fp16")
lavis/models/eva_vit.py:422
↓ 1 callersFunctioncreate_runner_config_validator
()
lavis/common/config.py:262
↓ 1 callersFunctiondefault_bpe
()
lavis/models/clip_models/tokenizer.py:25
↓ 1 callersMethoddefault_config_path
(cls, type="default")
lavis/datasets/builders/base_dataset_builder.py:86
↓ 1 callersFunctiondf_from_shelve
(chunk_size, func, dataset_name)
lavis/datasets/download_scripts/DownloadConceptualCaptions/download_data_cc3m.py:184
↓ 1 callersFunctiondf_from_shelve
(chunk_size, func, dataset_name)
lavis/datasets/download_scripts/DownloadConceptualCaptions/download_data_cc12m.py:184
↓ 1 callersFunctiondf_multiprocess
(df, processes, chunk_size, func, dataset_name)
lavis/datasets/download_scripts/DownloadConceptualCaptions/download_data_cc3m.py:40
↓ 1 callersFunctiondf_multiprocess
(df, processes, chunk_size, func, dataset_name)
lavis/datasets/download_scripts/DownloadConceptualCaptions/download_data_cc12m.py:40
↓ 1 callersMethoddispl_item
(self, index)
lavis/datasets/datasets/gqa_datasets.py:19
↓ 1 callersFunctiondownload_datasets
Download the Imagenet-R dataset archives and expand them in the folder provided as parameter
lavis/datasets/download_scripts/download_msrvtt.py:38
← previousnext →201–300 of 1,238, ranked by callers