A Diffusion Transformer model for 3D data from CogVideoX was introduced in CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer by Tsinghua University & ZhipuAI.
The model can be loaded with the following code snippet.
from diffusers import CogVideoXTransformer3DModel
vae = CogVideoXTransformer3DModel.from_pretrained("THUDM/CogVideoX-2b", subfolder="transformer", torch_dtype=torch.float16).to("cuda")
( num_attention_heads: int = 30 attention_head_dim: int = 64 in_channels: Optional = 16 out_channels: Optional = 16 flip_sin_to_cos: bool = True freq_shift: int = 0 time_embed_dim: int = 512 text_embed_dim: int = 4096 num_layers: int = 30 dropout: float = 0.0 attention_bias: bool = True sample_width: int = 90 sample_height: int = 60 sample_frames: int = 49 patch_size: int = 2 temporal_compression_ratio: int = 4 max_text_seq_length: int = 226 activation_fn: str = 'gelu-approximate' timestep_activation_fn: str = 'silu' norm_elementwise_affine: bool = True norm_eps: float = 1e-05 spatial_interpolation_scale: float = 1.875 temporal_interpolation_scale: float = 1.0 )
Parameters
int
, optional, defaults to 16) — The number of heads to use for multi-head attention. int
, optional, defaults to 88) — The number of channels in each head. int
, optional) —
The number of channels in the input. int
, optional) —
The number of channels in the output. int
, optional, defaults to 1) — The number of layers of Transformer blocks to use. float
, optional, defaults to 0.0) — The dropout probability to use. int
, optional) — The number of encoder_hidden_states
dimensions to use. bool
, optional) —
Configure if the TransformerBlocks
attention should contain a bias parameter. int
, optional) — The width of the latent images (specify if the input is discrete).
This is fixed during training since it is used to learn a number of position embeddings. int
, optional) —
The size of the patches to use in the patch embedding layer. str
, optional, defaults to "geglu"
) — Activation function to use in feed-forward. int
, optional) —
The number of diffusion steps used during training. Pass if at least one of the norm_layers is
AdaLayerNorm
. This is fixed during training since it is used to learn a number of embeddings that are
added to the hidden states. During inference, you can denoise for up to but not more steps than
num_embeds_ada_norm
. str
, optional, defaults to "layer_norm"
) —
The type of normalization to use. Options are "layer_norm"
or "ada_layer_norm"
. bool
, optional, defaults to True
) —
Whether or not to use elementwise affine in normalization layers. float
, optional, defaults to 1e-5) — The epsilon value to use in normalization layers. int
, optional) —
The number of channels in the caption embeddings. int
, optional) —
The number of frames in the video-like data. A Transformer model for video-like data in CogVideoX.
( sample: torch.Tensor )
Parameters
torch.Tensor
of shape (batch_size, num_channels, height, width)
or (batch size, num_vector_embeds - 1, num_latent_pixels)
if Transformer2DModel is discrete) —
The hidden states output conditioned on the encoder_hidden_states
input. If discrete, returns probability
distributions for the unnoised latent pixels. The output of Transformer2DModel.