Functions — github.com/PKU-YuanGroup/Helios

CodeHub MCPcopy Create free account

hub / github.com/PKU-YuanGroup/Helios / functions

Functions847 in github.com/PKU-YuanGroup/Helios

⨍ Functions 847◇ Types & classes 124

Functionclip_text_l14_336

( embed_dim=768, context_length=77, vocab_size=49408, transformer_width=768, transformer_h

eval/utils/third_party/ViCLIP/viclip_text.py:240

Functionclip_transform_Image

eval/utils/utils.py:32

Functioncollate_fn

helios/dataset/dataloader_history_latents_dist.py:529

Functioncollate_fn

helios/dataset/dataloader_mp4_dist.py:684

Functioncollate_fn

helios/dataset/dataloader_dmd.py:424

Functionconvert_GSB_csv_to_reward_data

Convert Good/Same/Bad csv data to reward data. Args: example (dict): A dataframe containing the GSB csv data. data_dir (str)

helios/videoalign/data.py:27

Copy current averaged parameters into given collection of parameters. Args: parameters: Iterable of `torch.nn.Parameter`

helios/utils/create_ema_zero3.py:287

Methodcurrent_timestep

helios/diffusers_version/pipeline_helios_diffusers.py:819

Methodcurrent_timestep

helios/pipelines/pipeline_helios_ode.py:914

Methodcurrent_timestep

helios/pipelines/pipeline_helios.py:849

Methoddo_classifier_free_guidance

helios/diffusers_version/pipeline_helios_diffusers.py:811

Methoddo_classifier_free_guidance

helios/pipelines/pipeline_helios_ode.py:906

Methoddo_classifier_free_guidance

helios/pipelines/pipeline_helios.py:841

Functionensure_contiguous

helios/modules/helios_kernels/tiled_linear.py:54

Functionevaluate_motion

(video_info, height=384, width=640)

eval/1_get_motion_amplitude.py:62

Functionextract_video_segment

尽可能保持原视频编码参数

eval/utils/utils.py:258

Functionflow_to_image

Expects a two dimensional flow image of shape. Args: flow_uv (np.ndarray): Flow UV image of shape [H,W,2] clip_flow (float, o

eval/utils/third_party/amt/utils/flow_utils.py:105

(self, img0, img1, embt, scale_factor=1.0, eval=False, **kwargs)

eval/utils/third_party/amt/networks/AMT-S.py:64

eval/utils/third_party/amt/networks/IFRNet.py:20

(self, f0, f1, embt)

eval/utils/third_party/amt/networks/IFRNet.py:35

(self, ft_, f0, f1, up_flow0, up_flow1)

eval/utils/third_party/amt/networks/IFRNet.py:50

(self, ft_, f0, f1, up_flow0, up_flow1)

eval/utils/third_party/amt/networks/IFRNet.py:65

(self, ft_, f0, f1, up_flow0, up_flow1)

eval/utils/third_party/amt/networks/IFRNet.py:80

(self, img0, img1, embt, scale_factor=1.0, eval=False, **kwargs)

eval/utils/third_party/amt/networks/IFRNet.py:97

(self, img0, img1, embt, scale_factor=1.0, eval=False, **kwargs)

eval/utils/third_party/amt/networks/AMT-G.py:65

(self, img0, img1, embt, scale_factor=1.0, eval=False, **kwargs)

eval/utils/third_party/amt/networks/AMT-L.py:65

eval/utils/third_party/amt/networks/blocks/feat_enc.py:50

eval/utils/third_party/amt/networks/blocks/feat_enc.py:102

eval/utils/third_party/amt/networks/blocks/feat_enc.py:161

eval/utils/third_party/amt/networks/blocks/feat_enc.py:235

eval/utils/third_party/amt/networks/blocks/feat_enc.py:311

eval/utils/third_party/amt/networks/blocks/ifrnet.py:40

eval/utils/third_party/amt/networks/blocks/ifrnet.py:70

(self, f0, f1, embt)

eval/utils/third_party/amt/networks/blocks/ifrnet.py:88

(self, ft_, f0, f1, flow0_in, flow1_in)

eval/utils/third_party/amt/networks/blocks/ifrnet.py:106

(self, ft_, f0, f1, flow0, flow1)

eval/utils/third_party/amt/networks/blocks/multi_flow.py:54

(self, net, flow, corr)

eval/utils/third_party/amt/networks/blocks/raft.py:64

(self, net, flow, corr)

eval/utils/third_party/amt/networks/blocks/raft.py:128

(self, tenInput)

eval/utils/third_party/amt/flow_generation/liteflownet/run.py:139

(self, tenOne, tenTwo, tenFeaturesOne, tenFeaturesTwo, tenFlow)

eval/utils/third_party/amt/flow_generation/liteflownet/run.py:208

(self, tenOne, tenTwo, tenFeaturesOne, tenFeaturesTwo, tenFlow)

eval/utils/third_party/amt/flow_generation/liteflownet/run.py:289

(self, tenOne, tenTwo, tenFeaturesOne, tenFeaturesTwo, tenFlow)

eval/utils/third_party/amt/flow_generation/liteflownet/run.py:391

(self, tenOne, tenTwo)

eval/utils/third_party/amt/flow_generation/liteflownet/run.py:467

(self, one, two, intStride)

eval/utils/third_party/amt/flow_generation/liteflownet/correlation/correlation.py:262

(self, tenOne, tenTwo, intStride)

eval/utils/third_party/amt/flow_generation/liteflownet/correlation/correlation.py:440

(self, **kwargs)

eval/utils/third_party/amt/losses/loss.py:19

(self, x: torch.Tensor)

eval/utils/third_party/ViCLIP/viclip_text.py:27

(self, x: torch.Tensor)

eval/utils/third_party/ViCLIP/viclip_text.py:34

(self, x: torch.Tensor)

eval/utils/third_party/ViCLIP/viclip_text.py:60

(self, x: torch.Tensor)

eval/utils/third_party/ViCLIP/viclip_text.py:75

eval/utils/third_party/ViCLIP/viclip_text.py:164

eval/utils/third_party/ViCLIP/viclip_vision.py:50

eval/utils/third_party/ViCLIP/viclip_vision.py:80

eval/utils/third_party/ViCLIP/viclip_vision.py:95

(self, x, masking_prob=0.0)

eval/utils/third_party/ViCLIP/viclip_vision.py:171

forward and calculate loss. Args: image (torch.Tensor): The input images. Shape: [B,T,C,H,W]. text (dict): TODO

eval/utils/third_party/ViCLIP/viclip.py:80

( self, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.Tensor] = N

helios/videoalign/trainer.py:30

helios/modules/transformer_helios.py:125

(self, hidden_states: torch.Tensor)

helios/modules/transformer_helios.py:148

(self, hidden_states: torch.Tensor, temb: torch.Tensor, original_context_length: int)

helios/modules/transformer_helios.py:163

( self, hidden_states: torch.Tensor, encoder_hidden_states: Optional[torch.Tensor] = N

helios/modules/transformer_helios.py:599

( self, timestep: torch.Tensor, encoder_hidden_states: Optional[torch.Tensor] = None,

helios/modules/transformer_helios.py:639

(self, frame_indices, height, width, device)

helios/modules/transformer_helios.py:697

( self, hidden_states: torch.Tensor, encoder_hidden_states: torch.Tensor, temb

helios/modules/transformer_helios.py:776

( self, hidden_states: torch.Tensor, timestep: torch.LongTensor, encoder_hidde

helios/modules/transformer_helios.py:1268

( ctx, fn: Callable, mlp_module: torch.nn.Module, x: torch.Tensor, sha

helios/modules/helios_kernels/tiled_linear.py:91

(self, hidden_states: torch.Tensor, *args, **kwargs)

helios/modules/helios_kernels/tiled_linear.py:265

Forward pass with tiled computation Args: hidden_states: [batch_size, seq_len, dim] or [seq_len, dim] Returns:

helios/modules/helios_kernels/tiled_linear.py:344

(ctx, X, W, b, eps)

helios/modules/helios_kernels/triton_norm.py:118

(ctx, X: torch.Tensor, W: torch.Tensor, eps: float, gemma: bool = False)

helios/modules/helios_kernels/triton_norm.py:331

(ctx, x, freqs_cis)

helios/modules/helios_kernels/triton_rope.py:66

(self, x, freqs_cis)

helios/modules/helios_kernels/triton_rope.py:270

( ctx, latent, discriminator, timestep, prompt_embeds, indices

helios/utils/utils_helios_post.py:311

(self, hidden_states: torch.Tensor, temb: torch.Tensor, original_context_length: int)

helios/diffusers_version/transformer_helios_diffusers.py:91

( self, hidden_states: torch.Tensor, encoder_hidden_states: torch.Tensor | None = None

helios/diffusers_version/transformer_helios_diffusers.py:271

( self, timestep: torch.Tensor, encoder_hidden_states: torch.Tensor | None = None,

helios/diffusers_version/transformer_helios_diffusers.py:307

(self, frame_indices, height, width, device)

helios/diffusers_version/transformer_helios_diffusers.py:355

( self, hidden_states: torch.Tensor, encoder_hidden_states: torch.Tensor, temb

helios/diffusers_version/transformer_helios_diffusers.py:426

( self, hidden_states: torch.Tensor, timestep: torch.LongTensor, encoder_hidde

helios/diffusers_version/transformer_helios_diffusers.py:661

Methodfuse_projections

helios/modules/transformer_helios.py:550

Methodfuse_projections

helios/diffusers_version/transformer_helios_diffusers.py:222

Functiongenerate_video

( mode: str, prompt: str, image_input, video_input, height: int, width: int, num_f

Functionget_global_rank

Find OMPI world rank without calling mpi functions :rtype: int

eval/utils/third_party/amt/utils/dist_utils.py:18

Functionget_local_rank

Find OMPI local rank without calling mpi functions :rtype: int

eval/utils/third_party/amt/utils/dist_utils.py:30

Functionget_master_ip

eval/utils/third_party/amt/utils/dist_utils.py:42

Functionget_memory_stats

helios/utils/utils_base.py:20

Methodget_num_layers

eval/utils/third_party/ViCLIP/viclip_vision.py:150

Functionget_peft_state_maybe_zero_3

(named_params, bias)

helios/videoalign/utils.py:110

Functionget_peft_state_non_lora_maybe_zero_3

(named_params, require_grad_only=True)

helios/videoalign/utils.py:135

Methodget_predict_label

(self, clip_feature, text_feats_tensor, top=5)

eval/utils/third_party/ViCLIP/viclip.py:213

Functionget_sigmas

(noise_scheduler, timesteps, n_dim=4, device="cuda", dtype=torch.float32)

helios/utils/utils_base.py:506

Methodget_text_encoder

get text encoder, used for text and cross-modal encoding

eval/utils/third_party/ViCLIP/viclip.py:191

Methodget_text_features

(self, input_text, tokenizer, text_feature_dict={})

eval/utils/third_party/ViCLIP/viclip.py:196

Functionget_timesteps

( num_inference_steps=50, denoising_strength=1, shift=1.0, num_train_timesteps=1000, sigma

helios/utils/utils_recycle_single.py:176

Methodget_vid_features

(self, input_frames)

eval/utils/third_party/ViCLIP/viclip.py:207

Methodguidance_scale

helios/diffusers_version/pipeline_helios_diffusers.py:807

Methodguidance_scale

helios/pipelines/pipeline_helios_ode.py:902

Methodguidance_scale

helios/pipelines/pipeline_helios.py:837

Functionhandle_idx_change

Handle index change - update steps and videos together

tools/gradio/comparison/gradio_compare_diff-video.py:354

Functioninterpolate_pos_embed_vit

(state_dict, new_model)

eval/utils/third_party/ViCLIP/viclip_vision.py:340

← previous next →701–800 of 847, ranked by callers