A Transformer model for image-like data from HiDream-I1.
The model can be loaded with the following code snippet.
from diffusers import HiDreamImageTransformer2DModel
transformer = HiDreamImageTransformer2DModel.from_pretrained("HiDream-ai/HiDream-I1-Full", subfolder="transformer", torch_dtype=torch.bfloat16)
( patch_size: typing.Optional[int] = None in_channels: int = 64 out_channels: typing.Optional[int] = None num_layers: int = 16 num_single_layers: int = 32 attention_head_dim: int = 128 num_attention_heads: int = 20 caption_channels: typing.List[int] = None text_emb_dim: int = 2048 num_routed_experts: int = 4 num_activated_experts: int = 2 axes_dims_rope: typing.Tuple[int, int] = (32, 32) max_resolution: typing.Tuple[int, int] = (128, 128) llama_layers: typing.List[int] = None )
( 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.