diffusers/pr_12652/en/api/models/chroma_transformer.md

ChromaTransformer2DModel

A modified flux Transformer model from Chroma

ChromaTransformer2DModel[[diffusers.ChromaTransformer2DModel]]

diffusers.ChromaTransformer2DModel[[diffusers.ChromaTransformer2DModel]]

Source

The Transformer model introduced in Flux, modified for Chroma.

Reference: https://huggingface.co/lodestones/Chroma1-HD

forwarddiffusers.ChromaTransformer2DModel.forwardhttps://github.com/huggingface/diffusers/blob/vr_12652/src/diffusers/models/transformers/transformer_chroma.py#L476[{"name": "hidden_states", "val": ": Tensor"}, {"name": "encoder_hidden_states", "val": ": Tensor = None"}, {"name": "timestep", "val": ": LongTensor = None"}, {"name": "img_ids", "val": ": Tensor = None"}, {"name": "txt_ids", "val": ": Tensor = None"}, {"name": "attention_mask", "val": ": Tensor = None"}, {"name": "joint_attention_kwargs", "val": ": dict[str, typing.Any] | None = None"}, {"name": "controlnet_block_samples", "val": " = None"}, {"name": "controlnet_single_block_samples", "val": " = None"}, {"name": "return_dict", "val": ": bool = True"}, {"name": "controlnet_blocks_repeat", "val": ": bool = False"}]- hidden_states (torch.Tensor of shape (batch_size, image_sequence_length, in_channels)) -- Input hidden_states.

• encoder_hidden_states (torch.Tensor of shape (batch_size, text_sequence_length, joint_attention_dim)) -- Conditional embeddings (embeddings computed from the input conditions such as prompts) to use.
• timestep ( torch.LongTensor) -- Used to indicate denoising step.
• block_controlnet_hidden_states -- (list of torch.Tensor): A list of tensors that if specified are added to the residuals of transformer blocks.
• joint_attention_kwargs (dict, optional) -- A kwargs dictionary that if specified is passed along to the AttentionProcessor as defined under self.processor in diffusers.models.attention_processor.
• return_dict (bool, optional, defaults to True) -- Whether or not to return a ~models.transformer_2d.Transformer2DModelOutput instead of a plain tuple.0If return_dict is True, an ~models.transformer_2d.Transformer2DModelOutput is returned, otherwise a tuple where the first element is the sample tensor.

The FluxTransformer2DModel forward method.

Parameters:

patch_size (int, defaults to 1) : Patch size to turn the input data into small patches.

in_channels (int, defaults to 64) : The number of channels in the input.

out_channels (int, optional, defaults to None) : The number of channels in the output. If not specified, it defaults to in_channels.

num_layers (int, defaults to 19) : The number of layers of dual stream DiT blocks to use.

num_single_layers (int, defaults to 38) : The number of layers of single stream DiT blocks to use.

attention_head_dim (int, defaults to 128) : The number of dimensions to use for each attention head.

num_attention_heads (int, defaults to 24) : The number of attention heads to use.

joint_attention_dim (int, defaults to 4096) : The number of dimensions to use for the joint attention (embedding/channel dimension of encoder_hidden_states).

axes_dims_rope (tuple[int], defaults to (16, 56, 56)) : The dimensions to use for the rotary positional embeddings.

Returns:

If return_dict is True, an ~models.transformer_2d.Transformer2DModelOutput is returned, otherwise a tuple where the first element is the sample tensor.