Functions — github.com/bytetriper/RAE

CodeHub MCPcopy Create free account

hub / github.com/bytetriper/RAE / functions

Functions315 in github.com/bytetriper/RAE

⨍ Functions 315◇ Types & classes 71

( self, dim: int, num_heads: int = 8, qkv_bias: bool = False, qk_norm:

src/stage2/models/model_utils.py:325

(self, hidden_size: int, embedding_size: int = 256, scale: float = 1.0)

src/stage2/models/model_utils.py:391

(self, num_classes, hidden_size, dropout_prob)

src/stage2/models/model_utils.py:415

src/eval/utils.py:28

(self, chn_in, chn_out=1, use_dropout=False)

src/eval/utils.py:38

(self, use_dropout=True)

src/eval/utils.py:85

src/eval/utils.py:129

(self, device: torch.device, **kwargs)

src/disc/discriminator.py:11

( self, input_size: int | Tuple[int, int], crop: int, num_windows: int,

src/disc/utils.py:27

(self, prob=1.0, cutout=0.2)

src/disc/diffaug.py:9

src/disc/lpips.py:16

(self, chn_in: int, chn_out: int = 1, use_dropout: bool = False)

src/disc/lpips.py:26

(self, requires_grad: bool = False, pretrained: bool = True)

src/disc/lpips.py:34

(self, in_features, hidden_features=None, out_features=None, fused_if_available=True)

src/disc/dinodisc.py:27

( self, block_idx, embed_dim=768, num_heads=12, flash_if_available=True, )

src/disc/dinodisc.py:59

(self, block_idx, embed_dim, num_heads, mlp_ratio, norm_eps)

src/disc/dinodisc.py:83

src/disc/dinodisc.py:97

(self, *args, **kwargs)

src/disc/dinodisc.py:108

(self, num_features: int, affine: bool = True, virtual_bs: int = 1, eps: float = 1e-6)

src/disc/dinodisc.py:114

(self, size: int)

src/disc/dinodisc.py:193

(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, norm_layer=None)

src/disc/dinodisc.py:272

( self, depth=12, key_depths=(2, 5, 8, 11), norm_eps=1e-6, patch_size=

src/disc/dinodisc.py:286

src/eval/utils.py:132

src/disc/utils.py:87

src/disc/dinodisc.py:204

src/disc/diffaug.py:18

Method_basic_init

src/stage2/models/DDT.py:283

Method_basic_init

src/stage2/models/lightningDiT.py:179

src/stage2/transport/integrators.py:109

(init, model, **model_kwargs)

src/stage2/transport/transport.py:398

Methodapprox_gelu

src/stage2/models/DDT.py:112

(x, t, model, **model_kwargs)

src/stage2/transport/transport.py:257

Functioncenter_crop_arr

Center cropping implementation from ADM. https://github.com/openai/guided-diffusion/blob/8fb3ad9197f16bbc40620447b2742e13458d2831/guided_diff

src/utils/train_utils.py:36

Methodcompute_alpha_t

Compute coefficient of x1

src/stage2/transport/path.py:179

Methodcompute_d_alpha_alpha_ratio_t

Special purposed function for computing numerical stabled d_alpha_t / alpha_t

src/stage2/transport/path.py:163

Methodcompute_d_alpha_alpha_ratio_t

Special purposed function for computing numerical stabled d_alpha_t / alpha_t

src/stage2/transport/path.py:191

Methodcompute_drift

Compute the drift term of the SDE

src/stage2/transport/path.py:168

Methodcompute_sigma_t

Compute coefficient of x1

src/stage2/transport/path.py:156

Methodcompute_sigma_t

Compute coefficient of x0

src/stage2/transport/path.py:185

(self, x: torch.Tensor)

src/stage1/rae.py:14

Functionequal_sampler

(step_idx: int)

src/sample_ddp.py:95

Methodextra_repr

src/disc/dinodisc.py:54

Methodextra_repr

src/disc/dinodisc.py:79

(self, x: torch.Tensor)

src/calculate_stat.py:86

(self, x: torch.Tensor)

src/stage1/rae.py:111

images is of shape (B, C, H, W) where B is batch size, C is number of channels, H and W are height and

src/stage1/encoders/siglip2.py:22

(self, x: torch.Tensor)

src/stage1/encoders/dinov2.py:35

images is of shape (B, C, H, W) where B is batch size, C is number of channels, H and W are height and width of the image

src/stage1/encoders/mae.py:21

(self, pixel_values, noise=None, interpolate_pos_encoding: bool = False)

src/stage1/decoders/decoder.py:275

(self, pixel_values, interpolate_pos_encoding: bool = False)

src/stage1/decoders/decoder.py:318

( self, hidden_states, head_mask: Optional[torch.Tensor] = None, output_attentions: bool = False )

src/stage1/decoders/decoder.py:358

( self, hidden_states, head_mask: Optional[torch.Tensor] = None, output_attentions: bool = False )

src/stage1/decoders/decoder.py:400

(self, hidden_states: torch.Tensor, input_tensor: torch.Tensor)

src/stage1/decoders/decoder.py:438

( self, hidden_states: torch.Tensor, head_mask: Optional[torch.Tensor] = None,

src/stage1/decoders/decoder.py:452

(self, hidden_states: torch.Tensor)

src/stage1/decoders/decoder.py:475

(self, hidden_states: torch.Tensor, input_tensor: torch.Tensor)

src/stage1/decoders/decoder.py:489

( self, hidden_states: torch.Tensor, head_mask: Optional[torch.Tensor] = None,

src/stage1/decoders/decoder.py:513

( self, hidden_states, output_attentions=False, output_hidden_states=False,

src/stage1/decoders/decoder.py:681

(self, *args: Any, **kwargs: Any)

src/stage2/__init__.py:8

(self, x, c, feat_rope=None)

src/stage2/models/DDT.py:137

src/stage2/models/DDT.py:174

(self, *args: Any, **kwargs: Any)

src/stage2/models/__init__.py:11

(self, x, c, feat_rope=None)

src/stage2/models/lightningDiT.py:80

src/stage2/models/lightningDiT.py:105

(self, t, start_index=0)

src/stage2/models/model_utils.py:130

src/stage2/models/model_utils.py:179

Returns: bias: Tensor of shape (1, num_heads, HW, HW) to be added to the raw attention logits before softmax.

src/stage2/models/model_utils.py:236

(self, x: torch.Tensor)

src/stage2/models/model_utils.py:268

Forward pass through the RMSNorm layer. Args: x (torch.Tensor): The input tensor. Returns: torch.Te

src/stage2/models/model_utils.py:306

(self, x: torch.Tensor, rope=None)

src/stage2/models/model_utils.py:355

src/stage2/models/model_utils.py:401

(self, labels, train, force_drop_ids=None)

src/stage2/models/model_utils.py:433

src/eval/utils.py:33

src/eval/utils.py:68

(self, input, target)

src/eval/utils.py:109

(self, tensor: torch.Tensor)

src/disc/lpips.py:21

(self, tensor: torch.Tensor)

src/disc/lpips.py:46

(self, input: torch.Tensor, target: torch.Tensor, reduction: str = "mean")

src/disc/lpips.py:93

src/disc/dinodisc.py:36

src/disc/dinodisc.py:69

src/disc/dinodisc.py:90

(self, x: torch.Tensor)

src/disc/dinodisc.py:102

(self, x: torch.Tensor)

src/disc/dinodisc.py:124

(self, x_in_pm1: torch.Tensor, grad_ckpt: bool = False)

src/disc/dinodisc.py:258

src/disc/dinodisc.py:280

(self, x, grad_ckpt=False)

src/disc/dinodisc.py:344

Methodforward_with_autoguidance

(self, *args: Any, **kwargs: Any)

src/stage2/__init__.py:10

Methodforward_with_autoguidance

Forward pass of LightningDiT, but also contain the forward pass for the additional model

src/stage2/models/DDT.py:391

Methodforward_with_autoguidance

(self, *args: Any, **kwargs: Any)

src/stage2/models/__init__.py:13

Methodforward_with_autoguidance

Forward pass of LightningDiT, but also contain the forward pass for the additional model

src/stage2/models/lightningDiT.py:284

Methodforward_with_cfg

(self, *args: Any, **kwargs: Any)

src/stage2/__init__.py:9

Methodforward_with_cfg

Forward pass of LightningDiT, but also batches the unconditional forward pass for classifier-free guidance.

src/stage2/models/DDT.py:364

Methodforward_with_cfg

(self, *args: Any, **kwargs: Any)

src/stage2/models/__init__.py:12

Methodforward_with_cfg

Forward pass of LightningDiT, but also batches the unconditional forward pass for classifier-free guidance.

src/stage2/models/lightningDiT.py:251

Methodget_noise_from_velocity

Wrapper function: transfrom velocity prediction model to denoiser Args: velocity: [batch_dim, ...] shaped tensor; velocity model o

src/stage2/transport/path.py:87

Methodget_velocity_from_score

Wrapper function: transfrom score prediction model to velocity Args: score: [batch_dim, ...] shaped tensor; score model output

src/stage2/transport/path.py:103

Functionhinge_d_loss

Hinge discriminator loss used by VQGAN.

src/disc/gan_loss.py:7

(stats, step=None)

src/utils/wandb_utils.py:88

Functionlog_image

(sample, step=None)

src/utils/wandb_utils.py:94

Functionlog_state

src/stage2/transport/utils.py:18

← previous next →201–300 of 315, ranked by callers