Functions — github.com/WillDreamer/Aurora

CodeHub MCPcopy Create free account

hub / github.com/WillDreamer/Aurora / functions

Functions226 in github.com/WillDreamer/Aurora

⨍ Functions 226◇ Types & classes 41

(self, *args, **kwargs)

image_root (string): Root directory of images (e.g. flickr30k/) ann_root (string): directory to store the annotation file

data/flickr30k_dataset.py:12

image_root (string): Root directory of images (e.g. flickr30k/) ann_root (string): directory to store the annotation file spl

data/flickr30k_dataset.py:53

(self, mean, std)

data/videovqa_dataset.py:18

(self, ann_root='MSRVTT-QA/', video_root = 'MSRVTT/videos/', train_files=[], split="train")

data/videovqa_dataset.py:30

image_root (string): Root directory of images (e.g. coco/images/) ann_root (string): directory to store the annotation file

data/coco_karpathy_dataset.py:9

image_root (string): Root directory of images (e.g. coco/images/) ann_root (string): directory to store the annotation file s

data/coco_karpathy_dataset.py:50

(self, mean, std)

data/video_dataset_didemo.py:35

image_root (string): Root directory of video ann_root (string): directory to store the annotation file

data/video_dataset_didemo.py:236

image_root (string): Root directory of video ann_root (string): directory to store the annotation file

data/video_dataset_didemo.py:306

(self, transform, ann_root, vqa_root, vg_root, train_files=[], split="train")

data/vqa_dataset.py:13

(self, mean, std)

data/video_dataset_msrvtt.py:18

image_root (string): Root directory of video ann_root (string): directory to store the annotation file

data/video_dataset_msrvtt.py:34

image_root (string): Root directory of video ann_root (string): directory to store the annotation file

data/video_dataset_msrvtt.py:116

(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0., R = 64)

(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0., dr

Args: img_size (int, tuple): input image size patch_size (int, tuple): patch size in_chans (int): number

Args: med_config (str): path for the mixture of encoder-decoder model's configuration file image_size (int): input im

CP/blip_vqa.py:10

Args: med_config (str): path for the mixture of encoder-decoder model's configuration file image_size (int): input im

CP/blip_retrieval.py:12

Args: med_config (str): path for the mixture of encoder-decoder model's configuration file image_size (int): input im

(self, config, is_cross_attention,R)

(self, config, iscross_attention)

(self, config, is_cross_attention=False,R = 64)

(self, config, adapter_config=None, adapters=None, device=None, R=None)

(self, config, add_pooling_layer=True, adapter_config=None, adapters=None, device=None,R=None)

(self, config, adapter_config=None, adapters=None, R = 64)

(self, N=2, M=10, isPIL=False, augs=[])

transform/randaugment.py:306

data/flickr30k_dataset.py:36

data/flickr30k_dataset.py:83

data/videovqa_dataset.py:47

data/coco_karpathy_dataset.py:33

data/coco_karpathy_dataset.py:82

data/video_dataset_didemo.py:284

data/video_dataset_didemo.py:347

data/vqa_dataset.py:40

data/video_dataset_msrvtt.py:65

data/video_dataset_msrvtt.py:140

Method_init_weights

Method_init_weights

Initialize the weights

Method_prune_heads

(self, heads_to_prune)

Method_reorder_cache

(self, past, beam_idx)

Functionautocontrast_func

same output as PIL.ImageOps.autocontrast

transform/randaugment.py:10

Functionblip_decoder

(pretrained='',**kwargs)

Functionblip_feature_extractor

(pretrained='',**kwargs)

Functionbrightness_func

same output as PIL.ImageEnhance.Contrast

transform/randaugment.py:116

Functioncoco_caption_eval

(coco_gt_root, results_file, split)

data/utils.py:93

Functioncolor_func

same output as PIL.ImageEnhance.Color

transform/randaugment.py:87

Functioncompute_acc

(logits, label, reduction='mean')

Functioncompute_n_params

(model, return_str=True)

Functioncontrast_func

same output as PIL.ImageEnhance.Contrast

transform/randaugment.py:103

Methodcreate_custom_forward

Methodcustom_forward

Functioncutout_func

(img, pad_size, replace=(0, 0, 0))

transform/randaugment.py:188

Functioncutout_level_to_args

(cutout_const, MAX_LEVEL, replace_value)

transform/randaugment.py:226

Functiondidemo_collate

data/__init__.py:102

Functionequalize_func

same output as PIL.ImageOps.equalize PIL's implementation is different from cv2.equalize

transform/randaugment.py:43

Methodfeed_forward_chunk

(self, attention_output)

(self, x, CP_U=None, CP_V=None, CP_C=None)

(self, x, register_hook=False, CP_U=None, CP_V=None, CP_C=None)

(self, x, register_hook=False, CP_U=None, CP_V=None, CP_C=None)

(self, x, register_blk=-1, CP_U=None, CP_V=None, CP_C=None)

(self, image, question, answer=None, n=None, weights=None, train=True, inference='rank', k_test=128)

CP/blip_vqa.py:47

(self, image, caption, alpha, idx)

CP/blip_retrieval.py:89

CP/blip_retrieval.py:372

(self, image, caption, mode)

(self, image, caption)

( self, input_ids=None, position_ids=None, inputs_embeds=None, past_key_values_length=0 )

( self, hidden_states, attention_mask=None, head_mask=None, encoder_hi

(self, hidden_states, input_tensor)

( self, hidden_states, attention_mask=None, head_mask=None, encoder_hi

(self, hidden_states)

(self, hidden_states, input_tensor)

( self, hidden_states, attention_mask=None, head_mask=None, encoder_hi

( self, hidden_states, attention_mask=None, head_mask=None, encoder_hi

(self, hidden_states)

(self, hidden_states)

(self, hidden_states)

(self, sequence_output)

r""" encoder_hidden_states (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`): Seq

r""" encoder_hidden_states (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`): Seq

Methodget_attention_map

Methodget_attention_map

Methodget_attn_gradients

Methodget_attn_gradients

Methodget_input_embeddings

Methodget_output_embeddings

Functionget_world_size

Methodglobal_avg

Functionidentity_func

transform/randaugment.py:6

Functioninit_distributed_mode

← previous next →101–200 of 226, ranked by callers