MCPcopy Create free account

hub / github.com/Francis-Rings/StableAnimator / functions

Functions480 in github.com/Francis-Rings/StableAnimator

⨍ Functions 480◇ Types & classes 92
↓ 37 callersMethodsplit
(self, indices)
animation/helper/eval/verification.py:47
↓ 30 callersMethodget
(self, rimg, landmark)
animation/helper/eval_ijbc.py:75
↓ 29 callersMethodfrom_pretrained
load pretrained pose-net weights
animation/modules/pose_net.py:66
↓ 18 callersMethodcheckpoint
(self, func, num_seg, x)
animation/helper/backbones/iresnet2060.py:145
↓ 16 callersMethod__init__
( self, in_channels: int, prev_output_channel: int, out_channels: int,
animation/modules/unet_3d_blocks.py:800
↓ 15 callersMethodstep
Predict the sample from the previous timestep by reversing the SDE. This function propagates the diffusion process from the learned
animation/pipelines/euler_discrete_pro.py:587
↓ 13 callersMethodupdate
(self, val, n=1)
animation/helper/utils/utils_logging.py:23
↓ 9 callersMethodencode
Encode a batch of images into latents. Args: x (`torch.Tensor`): Input batch of images. return_dict (`b
animation/modules/refined_vae.py:306
↓ 9 callersMethodset_attn_processor
r""" Sets the attention processor to use to compute attention. Parameters: processor (`dict` of `AttentionProcessor` or o
animation/modules/unet.py:284
↓ 8 callersMethoddecode
Decode a batch of images. Args: z (`torch.Tensor`): Input batch of latent vectors. return_dict (`bool`,
animation/modules/refined_vae.py:333
↓ 6 callersMethodrun
(self)
animation/helper/dataset.py:88
↓ 6 callersMethodsample
This functions will change the value of labels Parameters: ----------- labels: torch.Tensor
animation/helper/partial_fc_v2.py:79
↓ 5 callersMethod__init__
(self, in_features, hidden_features=None, out_features=None, act_layer=nn.ReLU6, drop=0.)
animation/helper/backbones/vit.py:7
↓ 5 callersMethod__init__
(self, embedding_size)
animation/helper/backbones/mobilefacenet.py:76
↓ 5 callersFunction_iresnet
(arch, block, layers, pretrained, progress, **kwargs)
animation/helper/backbones/iresnet.py:165
↓ 5 callersFunction_resize_with_antialiasing
(input, size, interpolation="bicubic", align_corners=True)
train.py:225
↓ 5 callersMethodbackward
(ctx, *grads)
animation/helper/partial_fc_v2.py:240
↓ 4 callersMethod__init__
(self, dali_iter)
animation/helper/dataset.py:269
↓ 4 callersMethod_make_layer
(self, block, planes, blocks, stride=1, dilate=False)
animation/helper/backbones/iresnet.py:122
↓ 4 callersMethod_make_layer
(self, block, planes, blocks, stride=1, dilate=False)
animation/helper/backbones/iresnet2060.py:119
↓ 4 callersMethodscale_model_input
Ensures interchangeability with schedulers that need to scale the denoising model input depending on the current timestep. Scales t
animation/pipelines/euler_discrete_pro.py:371
↓ 3 callersFunctioncalculate_accuracy
(threshold, dist, actual_issame)
animation/helper/eval/verification.py:109
↓ 3 callersMethodindex_for_timestep
(self, timestep, schedule_timesteps=None)
animation/pipelines/euler_discrete_pro.py:565
↓ 3 callersFunctionmultiplexing
(condition, true_case, false_case)
animation/helper/dataset.py:230
↓ 3 callersFunctionreshape_tensor
(x, heads)
animation/modules/id_encoder.py:6
↓ 2 callersMethod__init__
(self, cross_attention_dim=768, id_embeddings_dim=512, clip_embeddings_dim=1024, num_tokens=4)
animation/modules/id_encoder.py:110
↓ 2 callersMethod__init__
(self, s=64.0, margin=0.5)
animation/helper/losses.py:63
↓ 2 callersFunction_filter2d
(input, kernel)
train.py:278
↓ 2 callersFunction_filter2d
(input, kernel)
train_single.py:282
↓ 2 callersFunction_gaussian
(window_size: int, sigma)
train.py:303
↓ 2 callersFunction_gaussian
(window_size: int, sigma)
train_single.py:307
↓ 2 callersFunction_get_3rd_point
To calculate the affine matrix, three pairs of points are required. This function is used to get the 3rd point, given 2D points a & b. The 3r
DWPose/dwpose_utils/onnxpose.py:182
↓ 2 callersMethod_init_step_index
(self, timestep)
animation/pipelines/euler_discrete_pro.py:579
↓ 2 callersMethod_sigma_to_t
(self, sigma, log_sigmas)
animation/pipelines/euler_discrete_pro.py:516
↓ 2 callersFunctionalpha_bar_fn
(t)
animation/pipelines/euler_discrete_pro.py:163
↓ 2 callersFunctionalpha_blend_color
blend color according to point conf
DWPose/training_skeleton_extraction.py:14
↓ 2 callersFunctionalpha_blend_color
blend color according to point conf
DWPose/skeleton_extraction.py:11
↓ 2 callersFunctioncalculate_val_far
(threshold, dist, actual_issame)
animation/helper/eval/verification.py:165
↓ 2 callersMethodcheck
(self, track='cfat', test_img = None)
animation/helper/onnx_helper.py:22
↓ 2 callersFunctionconv3x3
3x3 convolution with padding
animation/helper/backbones/iresnet.py:8
↓ 2 callersFunctionconv3x3
3x3 convolution with padding
animation/helper/backbones/iresnet2060.py:10
↓ 2 callersFunctiondraw_pose
vis dwpose outputs Args: pose (List): DWposeDetector outputs in dwpose_detector.py H (int): height W (int): width
DWPose/skeleton_extraction.py:102
↓ 2 callersFunctiondraw_pose
(pose, H, W)
DWPose/dwpose_utils/__init__.py:15
↓ 2 callersFunctionface_embedding_extractor_get_model
(name, **kwargs)
animation/helper/backbones/__init__.py:5
↓ 2 callersMethodforward_db
(self, batch_data)
animation/helper/eval_ijbc.py:104
↓ 2 callersFunctionget_aligned_face
(face_loss_helper, frames, device)
animation/utils/utils.py:69
↓ 2 callersFunctionget_dist_info
()
animation/helper/utils/utils_distributed_sampler.py:34
↓ 2 callersFunctionload_images_from_folder
(folder)
train.py:157
↓ 2 callersFunctionload_images_from_folder
(folder)
train_single.py:159
↓ 2 callersFunctionlogsnr_schedule_cosine_shifted
(t, image_d, noise_d, logsnr_min, logsnr_max)
train.py:192
↓ 2 callersFunctionlogsnr_schedule_cosine_shifted
(t, image_d, noise_d, logsnr_min, logsnr_max)
train_single.py:194
↓ 2 callersMethodpreload
(self)
animation/helper/dataset.py:120
↓ 2 callersMethodreset
(self)
animation/helper/utils/utils_logging.py:17
↓ 2 callersFunctiontensor_to_vae_latent
(t, vae, scale=True)
train.py:368
↓ 2 callersFunctiontensor_to_vae_latent
(t, vae, scale=True)
train_single.py:372
↓ 1 callersFunctionFeedForward
(dim, mult=4)
animation/modules/id_encoder.py:60
↓ 1 callersMethod__init__
( self, hidden_size=None, cross_attention_dim=None, rank=4, netw
animation/modules/attention_processor.py:14
↓ 1 callersMethod__init__
( self, num_attention_heads: int = 16, attention_head_dim: int = 88, in_ch
animation/modules/transformer_temporal.py:71
↓ 1 callersMethod__init__
( self, num_attention_heads: int = 16, attention_head_dim: int = 88,
animation/modules/attention.py:57
↓ 1 callersMethod__init__
( self, in_channels: int = 4, out_channels: int = 3, block_out_channels: T
animation/modules/refined_vae.py:18
↓ 1 callersMethod__init__
Paramenters: ----------- embedding_size: int The dimension of embedding, required num_classes: int
animation/helper/partial_fc_v2.py:31
↓ 1 callersMethod__init__
(self)
animation/helper/lr_scheduler.py:48
↓ 1 callersMethod__init__
(self, block, layers, dropout=0, num_features=512, zero_init_residual=False,
animation/helper/backbones/iresnet.py:69
↓ 1 callersMethod__init__
(self, block, layers, dropout=0, num_features=512, zero_init_residual=False,
animation/helper/backbones/iresnet2060.py:67
↓ 1 callersMethod__iter__
(self)
animation/helper/dataset.py:279
↓ 1 callersFunction_append_dims
Appends dimensions to the end of a tensor until it has target_dims dimensions.
animation/pipelines/inference_pipeline_animation.py:23
↓ 1 callersFunction_append_dims
Appends dimensions to the end of a tensor until it has target_dims dimensions.
animation/pipelines/pipeline_animation.py:29
↓ 1 callersFunction_append_dims
Appends dimensions to the end of a tensor until it has target_dims dimensions.
animation/pipelines/validation_pipeline_animation.py:30
↓ 1 callersFunction_append_dims
Appends dimensions to the end of a tensor until it has target_dims dimensions.
animation/pipelines/inference_pipeline_animation_pro.py:30
↓ 1 callersFunction_compute_padding
Compute padding tuple.
train.py:255
↓ 1 callersFunction_compute_padding
Compute padding tuple.
train_single.py:259
↓ 1 callersMethod_convert_to_karras
Constructs the noise schedule of Karras et al. (2022).
animation/pipelines/euler_discrete_pro.py:540
↓ 1 callersMethod_encode_image
( self, image: PipelineImageInput, device: Union[str, torch.device],
animation/pipelines/inference_pipeline_animation.py:147
↓ 1 callersMethod_encode_image
( self, image: PipelineImageInput, device: Union[str, torch.device], num_vi
animation/pipelines/pipeline_animation.py:168
↓ 1 callersMethod_encode_image
( self, image: PipelineImageInput, device: Union[str, torch.device],
animation/pipelines/validation_pipeline_animation.py:175
↓ 1 callersMethod_encode_image
( self, image: PipelineImageInput, device: Union[str, torch.device],
animation/pipelines/inference_pipeline_animation_pro.py:150
↓ 1 callersMethod_encode_vae_image
( self, image: torch.Tensor, device: Union[str, torch.device],
animation/pipelines/inference_pipeline_animation.py:194
↓ 1 callersMethod_encode_vae_image
( self, image: torch.Tensor, device: Union[str, torch.device], num_videos_per_
animation/pipelines/pipeline_animation.py:215
↓ 1 callersMethod_encode_vae_image
( self, image: torch.Tensor, device: Union[str, torch.device],
animation/pipelines/validation_pipeline_animation.py:222
↓ 1 callersMethod_encode_vae_image
( self, image: torch.Tensor, device: Union[str, torch.device],
animation/pipelines/inference_pipeline_animation_pro.py:197
↓ 1 callersFunction_fix_aspect_ratio
Extend the scale to match the given aspect ratio. Args: scale (np.ndarray): The image scale (w, h) in shape (2, ) aspect_ratio (f
DWPose/dwpose_utils/onnxpose.py:149
↓ 1 callersFunction_gaussian_blur2d
(input, kernel_size, sigma)
train.py:320
↓ 1 callersFunction_gaussian_blur2d
(input, kernel_size, sigma)
train_single.py:324
↓ 1 callersFunction_get_add_time_ids
( fps, motion_bucket_id, noise_aug_strength, dtype, batch_size,
train.py:1235
↓ 1 callersFunction_get_add_time_ids
( fps, motion_bucket_id, noise_aug_strength, dtype,
train_single.py:1222
↓ 1 callersMethod_get_add_time_ids
( self, fps: int, motion_bucket_id: int, noise_aug_strengt
animation/pipelines/inference_pipeline_animation.py:217
↓ 1 callersMethod_get_add_time_ids
( self, fps: int, motion_bucket_id: int, noise_aug_strength: float, dt
animation/pipelines/pipeline_animation.py:238
↓ 1 callersMethod_get_add_time_ids
( self, fps: int, motion_bucket_id: int, noise_aug_strengt
animation/pipelines/validation_pipeline_animation.py:245
↓ 1 callersMethod_get_add_time_ids
( self, fps: int, motion_bucket_id: int, noise_aug_strengt
animation/pipelines/inference_pipeline_animation_pro.py:220
↓ 1 callersMethod_initialize_weights
Initialize weights with He. initialization and zero out the biases
animation/modules/pose_net.py:44
↓ 1 callersMethod_initialize_weights
(self)
animation/helper/backbones/mobilefacenet.py:120
↓ 1 callersFunction_iresnet
(arch, block, layers, pretrained, progress, **kwargs)
animation/helper/backbones/iresnet2060.py:168
↓ 1 callersMethod_read_txt_file_images
(self)
animation/dataset/animation_dataset.py:31
↓ 1 callersFunction_resize_with_antialiasing
(input, size, interpolation="bicubic", align_corners=True)
train_single.py:229
↓ 1 callersFunction_rotate_point
Rotate a point by an angle. Args: pt (np.ndarray): 2D point coordinates (x, y) in shape (2, ) angle_rad (float): rotation angle i
DWPose/dwpose_utils/onnxpose.py:167
↓ 1 callersMethod_set_gradient_checkpointing
(self, module, value=False)
animation/modules/unet.py:332
↓ 1 callersFunctionbbox_xyxy2cs
Transform the bbox format from (x,y,w,h) into (center, scale) Args: bbox (ndarray): Bounding box(es) in shape (4,) or (n, 4), formatted
DWPose/dwpose_utils/onnxpose.py:115
↓ 1 callersMethodbenchmark
(self, img)
animation/helper/onnx_helper.py:222
↓ 1 callersFunctionbetas_for_alpha_bar
Create a beta schedule that discretizes the given alpha_t_bar function, which defines the cumulative product of (1-beta) over time from t =
animation/pipelines/euler_discrete_pro.py:138
↓ 1 callersFunctioncalculate_roc
(thresholds, embeddings1, embeddings2, actual_issame,
animation/helper/eval/verification.py:54
next →1–100 of 480, ranked by callers