Pads images such that dimensions are divisible by 8

A residual block that can optionally change the number of channels. :param channels: the number of input channels. :param emb_channels: t

An attention block that allows spatial positions to attend to each other. Originally ported from here, but adapted to the N-d case. https

A downsampling layer with an optional convolution. :param channels: channels in the inputs and outputs. :param use_conv: a bool determini

Transformer block for image-like data. First, project the input (aka embedding) and reshape to b, t, d. Then apply standard transform

An upsampling layer with an optional convolution. :param channels: channels in the inputs and outputs. :param use_conv: a bool determinin

A module which performs QKV attention and splits in a different order.

Residual convolution module.

The full UNet model with attention and timestep embedding. :param in_channels: channels in the input Tensor. :param model_channels: base

This module implements SimpleFeaturePyramid in :paper:`vitdet`. It creates pyramid features built on top of the input feature map.

Feature fusion block.

This is a more standard version of the position embedding, very similar to the one used by the Attention is all you need paper, generalized

A module which performs QKV attention. Matches legacy QKVAttention + input/ouput heads shaping

Random contrast

Random contrast

flow propagation with self-attention on feature query: feature0, key: feature0, value: flow

self attention + cross attention + FFN

Adapted from CLIP: https://github.com/openai/CLIP/blob/main/clip/model.py

Corss attention conditioning An attention block that allows spatial positions to attend to each other. Originally ported from here, but a

Normalize image, with type tensor

Normalize image, with type tensor

Randomly vertically filps