Functions — github.com/UCSC-VLAA/OpenVision

CodeHub MCPcopy Create free account

hub / github.com/UCSC-VLAA/OpenVision / functions

Functions716 in github.com/UCSC-VLAA/OpenVision

⨍ Functions 716◇ Types & classes 90 ↳ Endpoints 1

(self, texts: Union[str, List[str]], context_length: Optional[int] = None)

src/convert_upload/open_clip/tokenizer.py:430

(self, texts: Union[str, List[str]], context_length: Optional[int] = None)

src/convert_upload/open_clip/tokenizer.py:501

(self, texts: Union[str, List[str]], context_length: Optional[int] = None)

src/convert_upload/open_clip/tokenizer.py:578

Args: img (PIL Image): Image to be cropped and resized. Returns: PIL Image: Resized, padded to at least targ

src/convert_upload/open_clip/transform.py:144

src/convert_upload/open_clip/transform.py:251

src/convert_upload/open_clip/transform.py:267

(self, orig_get_pp_fn)

src/helpers/registry.py:52

src/transforms/mixup.py:74

(self, *args, **kwargs)

src/evaluators/common.py:101

(self, image, text=None, image_latent=None, image_embs=None, output_labels=True, train=False, **kw)

src/models/two_towers.py:51

(self, q, k, v)

src/models/vit.py:55

src/models/vit.py:98

Applies Transformer MlpBlock module.

src/models/vit.py:189

src/models/vit.py:233

(self, x, deterministic=True)

src/models/vit.py:262

(self, x, deterministic=True)

src/models/vit.py:356

src/models/vit.py:410

(self, image, *, train=False)

src/models/vit.py:656

Performs the forward pass for image encoding and text decoding. Args: image_batch_pixels: A batch of raw input images, typically

src/models/openvision2_model.py:83

Applies Transformer MlpBlock module.

src/models/text_transformer.py:101

Applies multi-head dot product attention on the input data. Projects the inputs into multi-headed query, key, and value vectors, appl

src/models/text_transformer.py:146

(self, x, deterministic=True)

src/models/text_transformer.py:405

(self, x, deterministic=True)

src/models/text_transformer.py:531

(self, text, *, train=False, mask_ratio=0)

src/models/text_transformer.py:630

Applies multi-head dot product attention. Args: inputs_q: Input queries of shape `[B, L_q, D]`. inputs_kv: Keys/value

src/models/text_decoder_v2.py:84

(self, x: Array, u: Array, deterministic: bool = True)

src/models/text_decoder_v2.py:237

(self, x: Array, u: Array, deterministic: bool = True)

src/models/text_decoder_v2.py:310

(self, text_input: Array, context: Array, *, train: bool = False)

src/models/text_decoder_v2.py:387

Applies multi-head dot product attention on the input data. Projects the inputs into multi-headed query, key, and value vectors, applies dot-

src/models/common.py:254

Applies AddPositionEmbs module. Adds posemb to the inputs, supports single-position autoregressive mode. Args: inputs: input data [bat

src/models/common.py:627

(self, input, deterministic=None)

src/models/common.py:664

Applies multi-head dot product attention on the input data. Projects the inputs into multi-headed query, key, and value vectors, appl

src/models/text_decoder.py:63

(self, x, u, attn_mask=None, deterministic=True)

src/models/text_decoder.py:263

(self, x, u, deterministic=True)

src/models/text_decoder.py:351

(self, image_embeds, text_embeds, *, train=False)

src/models/text_decoder.py:437

( self, caption_loss_weight, clip_loss_weight, pad_id=0, # pa

src/convert_upload/open_clip/loss.py:135

( self, cache_labels=False, rank=0, world_size=1,

src/convert_upload/open_clip/loss.py:317

( self, embed_dim, multimodal_cfg: MultimodalCfg, text_cfg: CL

src/convert_upload/open_clip/coca_model.py:90

(self, spacial_dim: int, embed_dim: int, num_heads: int, output_dim: int = None)

src/convert_upload/open_clip/modified_resnet.py:59

(self, layers, output_dim, heads, image_size=224, width=64)

src/convert_upload/open_clip/modified_resnet.py:103

( self, bpe_path: str = default_bpe(), additional_special_tokens: Optional

src/convert_upload/open_clip/tokenizer.py:135

( self, tokenizer_name: str, context_length: Optional[int] = DEFAULT_CONTE

src/convert_upload/open_clip/tokenizer.py:407

( self, tokenizer_name: str, context_length: Optional[int] = 64, )

src/convert_upload/open_clip/tokenizer.py:474

(self, vocab_file, context_length=512, bos_token=1, eos_token=2, class_token=101, pad_token=0)

src/convert_upload/open_clip/tokenizer.py:525

( self, context_length: Optional[int] = DEFAULT_CONTEXT_LENGTH, clean: str

src/convert_upload/open_clip/tokenizer.py:555

( self, size, longest=0., interpolation=InterpolationMode.BICU

src/convert_upload/open_clip/transform.py:94

(self, size, fill=0)

src/convert_upload/open_clip/transform.py:219

(self, brightness=0., contrast=0., saturation=0., hue=0., p=0.8)

src/convert_upload/open_clip/transform.py:246

( self, model_name, embed_dim, image_size=224, poo

src/convert_upload/open_clip/timm_model.py:32

src/convert_upload/open_clip/hf_model.py:88

( self, model_name_or_path: str, output_dim: int, config: Pret

src/convert_upload/open_clip/hf_model.py:100

(self, dim, init_values=1e-5, inplace=False)

src/convert_upload/open_clip/transformer.py:40

(self, prob, exclude_first_token=True)

src/convert_upload/open_clip/transformer.py:54

( self, d_model: int, context_dim: int, n_head: int = 8,

src/convert_upload/open_clip/transformer.py:188

( self, d_model: int, n_head: int, mlp_ratio: float = 4.0,

src/convert_upload/open_clip/transformer.py:211

( self, d_model: int, n_head: int, mlp_ratio: float = 4.0,

src/convert_upload/open_clip/transformer.py:269

( self, width: int, layers: int, heads: int, mlp_r

src/convert_upload/open_clip/transformer.py:320

( self, width: int, layers: int, heads: int, mlp_r

src/convert_upload/open_clip/transformer.py:371

( self, image_size: int, patch_size: int, width: int,

src/convert_upload/open_clip/transformer.py:437

( self, context_length: int = 77, vocab_size: int = 49408, wid

src/convert_upload/open_clip/transformer.py:673

( self, width: int, layers: int, heads: int, conte

src/convert_upload/open_clip/transformer.py:821

( self, embed_dim: int, vision_cfg: CLIPVisionCfg, text_cfg: C

src/convert_upload/open_clip/model.py:323

src/helpers/utils.py:281

(self, xid=-1, wid=-1, workdir=None, config=None)

src/helpers/utils.py:839

(self, indefault="image", outdefault="image", with_data=False)

src/helpers/registry.py:47

Applies Mixup and/or Cutmix to a batch of images. Args: mixup_alpha (float, optional): For drawing a random lambda (`lam`) from a be

src/transforms/mixup.py:36

Applies RandomErasing to a single image. Args: probability: Probability of augmenting the image. Defaults to `0.25`. min_area: Minimum

src/transforms/random_erasing.py:106

(self, name, split, data_dir=None, skip_decode=("image",))

src/datasets/tfds.py:29

Initializes a new zero-shot image/text retrieval evaluator. See `prepare_datasets()` for details on how the dataset is pre-processed. Args:

src/evaluators/proj/image_text/retrieval.py:157

(self, predict_fn, data, pp_fn, batch_size, use_global_batch, cache_final=True,

src/evaluators/proj/image_text/contrastive.py:76

Initializes a new zero-shot classification evaluator. See `prepare_datasets()` for details on how the dataset is pre-processed. Args

src/evaluators/proj/image_text/discriminative_classifier.py:181

Method__post_init__

src/convert_upload/open_clip/transform.py:26

src/convert_upload/open_clip/transform.py:160

src/convert_upload/open_clip/transform.py:234

src/helpers/utils.py:612

Function_broadcast

(leaf, subtree)

src/helpers/utils.py:502

Function_chunk_attention_bias

(query_chunk_size, key_chunk_size, bias, deterministic, attn_dropout, attn_pdrop, causal,

src/models/bpt.py:164

Function_clean_canonicalize

src/convert_upload/open_clip/tokenizer.py:79

Function_clean_lower

src/convert_upload/open_clip/tokenizer.py:84

Function_clean_whitespace

src/convert_upload/open_clip/tokenizer.py:89

Function_color_jitter

src/transforms/ops_image.py:361

Function_color_jitter_timm

src/transforms/ops_image.py:388

Function_concat

src/transforms/ops_general.py:207

Function_convert_ln

src/convert_upload/open_clip/factory.py:300

Function_convert_to_rgb

src/convert_upload/open_clip/transform.py:238

Function_convert_weights

src/convert_upload/open_clip/model.py:401

src/transforms/ops_general.py:165

Method_create_block

src/convert_upload/open_clip/transformer.py:393

src/transforms/ops_image.py:259

Function_decode

src/transforms/ops_image.py:46

Function_do_the_mapping

Map string to number.

src/transforms/ops_general.py:93

src/transforms/ops_general.py:155

Function_drop_no_imagenet

src/configs/clip_common.py:59

Function_drop_no_real_label

src/configs/clip_common.py:55

Function_enhance_increasing_level_to_arg

src/transforms/autoaugment.py:499

Function_enhance_level_to_arg

src/transforms/autoaugment.py:495

Function_eval_fn

(params, images, labels, mask)

src/evaluators/proj/image_text/contrastive.py:55

src/helpers/utils.py:582

Function_get_dataset_info

src/evaluators/proj/image_text/discriminative_classifier.py:65

Method_global_pool

(self, x: jnp.ndarray)

src/models/vit.py:638

← previous next →301–400 of 716, ranked by callers