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

A sequential module that passes timestep embeddings to the children that support it as an extra input.

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

Uses the CLIP image encoder. Not actually frozen... If you want that set cond_stage_trainable=False in cfg

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

Uses the CLIP transformer encoder for text (from huggingface)

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 named tuple describing a ResNet block.

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

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

Uses a pretrained BERT tokenizer by huggingface. Vocab size: 30522 (?)

Chains list of learning rate schedulers. It takes a list of chainable learning rate schedulers and performs consecutive step() functions belong to

Uses the T5 transformer encoder for text

to match AttnBlock usage

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

Receives the list of schedulers that is expected to be called sequentially during optimization process and milestone points that provides exact in

Uses the BERT tokenizr model and add some transformer encoder layers

Two ways to print to console: 1. self.print: correctly handle progress bar 2. rank_zero_info: use the logging module