MCPcopy Create free account

hub / github.com/ali-vilab/TeaCache / functions

Functions1,040 in github.com/ali-vilab/TeaCache

⨍ Functions 1,040◇ Types & classes 227
Methodcreate_custom_forward
(module)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:478
Methodcreate_custom_forward
(module)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:579
Methodcreate_custom_forward
(module)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:701
Methodcreate_custom_forward
(module)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:842
Methodcreate_forward
(*inputs)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:391
Methodcreate_forward
(*inputs)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:479
Methodcreate_forward
(*inputs)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:580
Functioncustom_forward
(*inputs)
TeaCache4ConsisID/teacache_sample_video.py:108
Functioncustom_forward
(*inputs)
TeaCache4LTX-Video/teacache_ltx.py:94
Functioncustom_forward
(*inputs)
TeaCache4TangoFlux/teacache_tango_flux.py:114
Functioncustom_forward
(*inputs)
eval/teacache/experiments/cogvideox.py:103
Functioncustom_forward
(*inputs)
TeaCache4FLUX/teacache_flux.py:135
Functioncustom_forward
(*inputs)
TeaCache4Mochi/teacache_mochi.py:92
Functioncustom_forward
(*inputs)
TeaCache4CogVideoX1.5/teacache_sample_video.py:99
Methodcustom_forward
(*inputs)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:1906
Methodcustom_forward
(*inputs)
videosys/models/transformers/cogvideox_transformer_3d.py:541
Methodcustom_forward
(*inputs)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:702
Methodcustom_forward
(*inputs)
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:843
Methoddecode
(self, x)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan.py:812
Methoddecode
(self, x, **kwargs)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:522
Methoddecode
(self, x, **kwargs)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:570
Methoddevice
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:587
Methoddevice
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:723
Methoddictionary_lookup
(self, encodings)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan_v110.py:1376
Methoddictionary_lookup
(self, encodings)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan.py:1373
Methoddisable_slicing
r""" Disable sliced VAE decoding. If `enable_slicing` was previously enabled, this method will go back to computing decoding in one st
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:1064
Methoddisable_tiling
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan_v110.py:776
Methoddisable_tiling
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan_v120.py:1048
Methoddisable_tiling
r""" Disable tiled VAE decoding. If `enable_tiling` was previously enabled, this method will go back to computing decoding in one step
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:1050
Methoddisable_tiling
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan.py:776
Methoddo_classifier_free_guidance
(self)
videosys/pipelines/vchitect/pipeline_vchitect.py:695
Methoddtype
(self)
videosys/models/modules/embeddings.py:179
Methoddtype
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan_v110.py:818
Methoddtype
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan_v120.py:1138
Methoddtype
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan.py:818
Methoddtype
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:554
Methoddtype
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:591
Methoddtype
(self)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:727
Functionempty_cache
(func)
videosys/utils/test.py:6
Methodenable_forward_chunking
Sets the attention processor to use [feed forward chunking](https://huggingface.co/blog/reformer#2-chunked-feed-forward-layers).
videosys/models/transformers/vchitect_transformer_3d.py:350
Methodenable_parallel
(self, dp_size, sp_size, enable_cp)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:1716
Methodenable_parallel
(self, dp_size, sp_size, enable_cp)
videosys/models/transformers/vchitect_transformer_3d.py:326
Methodenable_parallel
(self, dp_size, sp_size, enable_cp)
videosys/models/transformers/cogvideox_transformer_3d.py:462
Methodenable_parallel
(self, dp_size, sp_size, enable_cp)
videosys/models/transformers/open_sora_plan_v110_transformer_3d.py:2364
Methodenable_parallel
(self, dp_size, sp_size, enable_cp)
videosys/models/transformers/open_sora_transformer_3d.py:443
Methodenable_tiling
r""" Enable tiled VAE decoding. When this option is enabled, the VAE will split the input tensor into tiles to compute decoding and en
videosys/models/autoencoders/autoencoder_kl_cogvideox.py:1014
Methodencode
(self, x)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan_v110.py:807
Methodencode
(self, x)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan_v120.py:1124
Methodencode
(self, x)
videosys/models/autoencoders/autoencoder_kl_open_sora_plan.py:807
Methodencode
(self, x)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:503
Methodencode
(self, x)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:567
Methodexecute_method_async
(self, method: str, *args, **kwargs)
videosys/core/mp_utils.py:256
Functionexists
(v)
videosys/models/autoencoders/autoencoder_kl_open_sora.py:85
Methodfn_recursive_add_processors
( name: str, module: torch.nn.Module, processors: Dict[str, VchitectAttnProcessor] )
videosys/models/transformers/vchitect_transformer_3d.py:390
Methodfn_recursive_attn_processor
(name: str, module: torch.nn.Module, processor)
videosys/models/transformers/vchitect_transformer_3d.py:428
Methodfn_recursive_feed_forward
(module: torch.nn.Module, chunk_size: int, dim: int)
videosys/models/transformers/vchitect_transformer_3d.py:369
Methodforward
Returns: outputs: Tensor handle: Optional[Work], if overlap is True
videosys/core/comm.py:36
Methodforward
Returns: outputs: Tensor handle: Optional[Work], if overlap is True
videosys/core/comm.py:160
Methodforward
(ctx, input_, process_group, scatter_dim, gather_dim)
videosys/core/comm.py:226
Methodforward
(ctx, input_, process_group, dim, grad_scale, pad)
videosys/core/comm.py:310
Methodforward
(ctx, input_, process_group, dim, grad_scale, pad)
videosys/core/comm.py:342
Methodforward
(self, hidden_states)
videosys/core/shardformer/t5/modeling.py:14
Methodforward
(self, x: torch.Tensor)
videosys/models/modules/downsampling.py:40
Methodforward
(self, inputs: torch.Tensor)
videosys/models/modules/upsampling.py:39
Methodforward
(self, x: torch.Tensor)
videosys/models/modules/attentions.py:56
Methodforward
(self, x, cond, mask=None)
videosys/models/modules/attentions.py:119
Methodforward
r""" The forward method of the `Attention` class. Args: hidden_states (`torch.Tensor`): The hidden states
videosys/models/modules/attentions.py:451
Methodforward
r""" Args: text_embeds (`torch.Tensor`): Input text embeddings. Expected shape: (batch_size, seq_length, embedding
videosys/models/modules/embeddings.py:31
Methodforward
Forward function.
videosys/models/modules/embeddings.py:85
Methodforward
(self, t, dtype)
videosys/models/modules/embeddings.py:141
Methodforward
(self, s, bs)
videosys/models/modules/embeddings.py:164
Methodforward
(self, caption, train, force_drop_ids=None)
videosys/models/modules/embeddings.py:221
Methodforward
( self, x: torch.Tensor, h: int, w: int, scale: Optional[float] = 1.0,
videosys/models/modules/embeddings.py:272
Methodforward
(self, hidden_states)
videosys/models/modules/normalization.py:17
Methodforward
( self, hidden_states: torch.Tensor, encoder_hidden_states: torch.Tensor, temb: torch.Tensor )
videosys/models/modules/normalization.py:40
Methodforward
( self, x: torch.Tensor, timestep: Optional[torch.Tensor] = None, temb: Optional[torch.Tensor] = None
videosys/models/modules/normalization.py:85
Methodforward
(self, f: torch.Tensor, zq: torch.Tensor)
videosys/models/modules/normalization.py:127
Methodforward
input: * tokens: batch_size x nheads x ntokens x dim * positions: batch_size x ntokens x 3 (t, y and x position of ea
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:97
Methodforward
(self, latent, num_frames)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:300
Methodforward
(self, latent, num_frames)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:431
Methodforward
(self, latent, num_frames)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:567
Methodforward
(self, x, attention_mask, t, h, w)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:743
Methodforward
(self, x, attention_mask, t, h, w)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:801
Methodforward
(self, x, t, h, w)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:1009
Methodforward
(self, x, t, h, w)
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:1078
Methodforward
( self, hidden_states: torch.FloatTensor, attention_mask: Optional[torch.FloatTensor]
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:1322
Methodforward
The [`Transformer2DModel`] forward method. Args: hidden_states (`torch.LongTensor` of shape `(batch size, num latent pix
videosys/models/transformers/open_sora_plan_v120_transformer_3d.py:1734
Methodforward
( self, hidden_states: torch.FloatTensor, encoder_hidden_states: torch.FloatTensor,
videosys/models/transformers/vchitect_transformer_3d.py:114
Methodforward
(self, hidden_states: torch.Tensor, *args, **kwargs)
videosys/models/transformers/vchitect_transformer_3d.py:228
Methodforward
The [`VchitectXLTransformerModel`] forward method. Args: hidden_states (`torch.FloatTensor` of shape `(batch size, chann
videosys/models/transformers/vchitect_transformer_3d.py:489
Methodforward
( self, hidden_states: torch.Tensor, encoder_hidden_states: torch.Tensor, temb
videosys/models/transformers/cogvideox_transformer_3d.py:268
Methodforward
( self, hidden_states: torch.Tensor, encoder_hidden_states: torch.Tensor, time
videosys/models/transformers/cogvideox_transformer_3d.py:479
Methodforward
(self, x: torch.Tensor, objs: torch.Tensor)
videosys/models/transformers/latte_transformer_3d.py:79
Methodforward
(self, hidden_states: torch.Tensor, scale: float = 1.0)
videosys/models/transformers/latte_transformer_3d.py:139
Methodforward
( self, hidden_states: torch.FloatTensor, attention_mask: Optional[torch.FloatTensor]
videosys/models/transformers/latte_transformer_3d.py:357
Methodforward
( self, hidden_states: torch.FloatTensor, attention_mask: Optional[torch.FloatTensor]
videosys/models/transformers/latte_transformer_3d.py:680
Methodforward
( self, timestep: torch.Tensor, added_cond_kwargs: Dict[str, torch.Tensor] = None,
videosys/models/transformers/latte_transformer_3d.py:867
Methodforward
The [`Transformer2DModel`] forward method. Args: hidden_states (`torch.LongTensor` of shape `(batch size, num latent pix
videosys/models/transformers/latte_transformer_3d.py:1144
Methodforward
(self, tokens, positions)
videosys/models/transformers/open_sora_plan_v110_transformer_3d.py:190
Methodforward
input: * tokens: batch_size x nheads x ntokens x dim * positions: batch_size x ntokens (t position of each token)
videosys/models/transformers/open_sora_plan_v110_transformer_3d.py:229
← previousnext →701–800 of 1,040, ranked by callers