Diffusers 文档
CogView3PlusTransformer2DModel
并获得增强的文档体验
开始使用
CogView3PlusTransformer2DModel
来自 CogView3Plus 的 2D 数据扩散 Transformer 模型在清华大学和智谱AI 的 CogView3: Finer and Faster Text-to-Image Generation via Relay Diffusion 中被介绍。
该模型可以通过以下代码片段加载。
from diffusers import CogView3PlusTransformer2DModel
transformer = CogView3PlusTransformer2DModel.from_pretrained("THUDM/CogView3Plus-3b", subfolder="transformer", torch_dtype=torch.bfloat16).to("cuda")CogView3PlusTransformer2DModel
class diffusers.CogView3PlusTransformer2DModel
< 来源 >( patch_size: int = 2 in_channels: int = 16 num_layers: int = 30 attention_head_dim: int = 40 num_attention_heads: int = 64 out_channels: int = 16 text_embed_dim: int = 4096 time_embed_dim: int = 512 condition_dim: int = 256 pos_embed_max_size: int = 128 sample_size: int = 128 )
参数
- patch_size (
int, 默认为2) — 在补丁嵌入层中使用的补丁大小。 - in_channels (
int, 默认为16) — 输入中的通道数。 - num_layers (
int, 默认为30) — 要使用的 Transformer 块层数。 - attention_head_dim (
int, 默认为40) — 每个头的通道数。 - num_attention_heads (
int, 默认为64) — 多头注意力使用的头数。 - out_channels (
int, 默认为16) — 输出中的通道数。 - text_embed_dim (
int, 默认为4096) — 文本编码器文本嵌入的输入维度。 - time_embed_dim (
int, 默认为512) — 时间步嵌入的输出维度。 - condition_dim (
int, 默认为256) — 输入 SDXL 风格分辨率条件（original_size、target_size、crop_coords）的嵌入维度。 - pos_embed_max_size (
int, 默认为128) — 位置嵌入的最大分辨率，从中获取形状为H x W的切片并添加到输入打补丁的潜变量中，其中H和W分别是潜在变量的高度和宽度。值为 128 意味着图像生成的最大支持高度和宽度为128 * vae_scale_factor * patch_size => 128 * 8 * 2 => 2048。 - sample_size (
int, 默认为128) — 输入潜变量的基础分辨率。如果在生成期间未提供高度/宽度，则此值用于确定分辨率为sample_size * vae_scale_factor => 128 * 8 => 1024
在 CogView3: Finer and Faster Text-to-Image Generation via Relay Diffusion 中引入的 Transformer 模型。
forward
< 来源 >( hidden_states: Tensor encoder_hidden_states: Tensor timestep: LongTensor original_size: Tensor target_size: Tensor crop_coords: Tensor return_dict: bool = True ) → torch.Tensor 或 ~models.transformer_2d.Transformer2DModelOutput
参数
- hidden_states (
torch.Tensor) — 形状为(批大小, 通道, 高度, 宽度)的输入hidden_states。 - encoder_hidden_states (
torch.Tensor) — 形状为(批大小, 序列长度, text_embed_dim)的条件嵌入（从提示等输入条件计算的嵌入） - timestep (
torch.LongTensor) — 用于指示去噪步骤。 - original_size (
torch.Tensor) — CogView3 使用 SDXL 风格的微条件来表示原始图像大小，如 https://huggingface.ac.cn/papers/2307.01952 第 2.2 节所述。 - target_size (
torch.Tensor) — CogView3 使用 SDXL 风格的微条件来表示目标图像大小，如 https://huggingface.ac.cn/papers/2307.01952 第 2.2 节所述。 - crop_coords (
torch.Tensor) — CogView3 使用 SDXL 风格的微条件来表示裁剪坐标，如 https://huggingface.ac.cn/papers/2307.01952 第 2.2 节所述。 - return_dict (
bool, 可选, 默认为True) — 是否返回~models.transformer_2d.Transformer2DModelOutput而不是普通元组。
返回
torch.Tensor 或 ~models.transformer_2d.Transformer2DModelOutput
使用提供的输入作为条件去噪后的潜在变量。
CogView3PlusTransformer2DModel 的 forward 方法。
设置注意力处理器
< 来源 >( processor: typing.Union[diffusers.models.attention_processor.AttnProcessor, diffusers.models.attention_processor.CustomDiffusionAttnProcessor, diffusers.models.attention_processor.AttnAddedKVProcessor, diffusers.models.attention_processor.AttnAddedKVProcessor2_0, diffusers.models.attention_processor.JointAttnProcessor2_0, diffusers.models.attention_processor.PAGJointAttnProcessor2_0, diffusers.models.attention_processor.PAGCFGJointAttnProcessor2_0, diffusers.models.attention_processor.FusedJointAttnProcessor2_0, diffusers.models.attention_processor.AllegroAttnProcessor2_0, diffusers.models.attention_processor.AuraFlowAttnProcessor2_0, diffusers.models.attention_processor.FusedAuraFlowAttnProcessor2_0, diffusers.models.attention_processor.FluxAttnProcessor2_0, diffusers.models.attention_processor.FluxAttnProcessor2_0_NPU, diffusers.models.attention_processor.FusedFluxAttnProcessor2_0, diffusers.models.attention_processor.FusedFluxAttnProcessor2_0_NPU, diffusers.models.attention_processor.CogVideoXAttnProcessor2_0, diffusers.models.attention_processor.FusedCogVideoXAttnProcessor2_0, diffusers.models.attention_processor.XFormersAttnAddedKVProcessor, diffusers.models.attention_processor.XFormersAttnProcessor, diffusers.models.attention_processor.XLAFlashAttnProcessor2_0, diffusers.models.attention_processor.AttnProcessorNPU, diffusers.models.attention_processor.AttnProcessor2_0, diffusers.models.attention_processor.MochiVaeAttnProcessor2_0, diffusers.models.attention_processor.MochiAttnProcessor2_0, diffusers.models.attention_processor.StableAudioAttnProcessor2_0, diffusers.models.attention_processor.HunyuanAttnProcessor2_0, diffusers.models.attention_processor.FusedHunyuanAttnProcessor2_0, diffusers.models.attention_processor.PAGHunyuanAttnProcessor2_0, diffusers.models.attention_processor.PAGCFGHunyuanAttnProcessor2_0, diffusers.models.attention_processor.LuminaAttnProcessor2_0, diffusers.models.attention_processor.FusedAttnProcessor2_0, diffusers.models.attention_processor.CustomDiffusionXFormersAttnProcessor, diffusers.models.attention_processor.CustomDiffusionAttnProcessor2_0, diffusers.models.attention_processor.SlicedAttnProcessor, diffusers.models.attention_processor.SlicedAttnAddedKVProcessor, diffusers.models.attention_processor.SanaLinearAttnProcessor2_0, diffusers.models.attention_processor.PAGCFGSanaLinearAttnProcessor2_0, diffusers.models.attention_processor.PAGIdentitySanaLinearAttnProcessor2_0, diffusers.models.attention_processor.SanaMultiscaleLinearAttention, diffusers.models.attention_processor.SanaMultiscaleAttnProcessor2_0, diffusers.models.attention_processor.SanaMultiscaleAttentionProjection, diffusers.models.attention_processor.IPAdapterAttnProcessor, diffusers.models.attention_processor.IPAdapterAttnProcessor2_0, diffusers.models.attention_processor.IPAdapterXFormersAttnProcessor, diffusers.models.attention_processor.SD3IPAdapterJointAttnProcessor2_0, diffusers.models.attention_processor.PAGIdentitySelfAttnProcessor2_0, diffusers.models.attention_processor.PAGCFGIdentitySelfAttnProcessor2_0, diffusers.models.attention_processor.LoRAAttnProcessor, diffusers.models.attention_processor.LoRAAttnProcessor2_0, diffusers.models.attention_processor.LoRAXFormersAttnProcessor, diffusers.models.attention_processor.LoRAAttnAddedKVProcessor, typing.Dict[str, typing.Union[diffusers.models.attention_processor.AttnProcessor, diffusers.models.attention_processor.CustomDiffusionAttnProcessor, diffusers.models.attention_processor.AttnAddedKVProcessor, diffusers.models.attention_processor.AttnAddedKVProcessor2_0, diffusers.models.attention_processor.JointAttnProcessor2_0, diffusers.models.attention_processor.PAGJointAttnProcessor2_0, diffusers.models.attention_processor.PAGCFGJointAttnProcessor2_0, diffusers.models.attention_processor.FusedJointAttnProcessor2_0, diffusers.models.attention_processor.AllegroAttnProcessor2_0, diffusers.models.attention_processor.AuraFlowAttnProcessor2_0, diffusers.models.attention_processor.FusedAuraFlowAttnProcessor2_0, diffusers.models.attention_processor.FluxAttnProcessor2_0, diffusers.models.attention_processor.FluxAttnProcessor2_0_NPU, diffusers.models.attention_processor.FusedFluxAttnProcessor2_0, diffusers.models.attention_processor.FusedFluxAttnProcessor2_0_NPU, diffusers.models.attention_processor.CogVideoXAttnProcessor2_0, diffusers.models.attention_processor.FusedCogVideoXAttnProcessor2_0, diffusers.models.attention_processor.XFormersAttnAddedKVProcessor, diffusers.models.attention_processor.XFormersAttnProcessor, diffusers.models.attention_processor.XLAFlashAttnProcessor2_0, diffusers.models.attention_processor.AttnProcessorNPU, diffusers.models.attention_processor.AttnProcessor2_0, diffusers.models.attention_processor.MochiVaeAttnProcessor2_0, diffusers.models.attention_processor.MochiAttnProcessor2_0, diffusers.models.attention_processor.StableAudioAttnProcessor2_0, diffusers.models.attention_processor.HunyuanAttnProcessor2_0, diffusers.models.attention_processor.FusedHunyuanAttnProcessor2_0, diffusers.models.attention_processor.PAGHunyuanAttnProcessor2_0, diffusers.models.attention_processor.PAGCFGHunyuanAttnProcessor2_0, diffusers.models.attention_processor.LuminaAttnProcessor2_0, diffusers.models.attention_processor.FusedAttnProcessor2_0, diffusers.models.attention_processor.CustomDiffusionXFormersAttnProcessor, diffusers.models.attention_processor.CustomDiffusionAttnProcessor2_0, diffusers.models.attention_processor.SlicedAttnProcessor, diffusers.models.attention_processor.SlicedAttnAddedKVProcessor, diffusers.models.attention_processor.SanaLinearAttnProcessor2_0, diffusers.models.attention_processor.PAGCFGSanaLinearAttnProcessor2_0, diffusers.models.attention_processor.PAGIdentitySanaLinearAttnProcessor2_0, diffusers.models.attention_processor.SanaMultiscaleLinearAttention, diffusers.models.attention_processor.SanaMultiscaleAttnProcessor2_0, diffusers.models.attention_processor.SanaMultiscaleAttentionProjection, diffusers.models.attention_processor.IPAdapterAttnProcessor, diffusers.models.attention_processor.IPAdapterAttnProcessor2_0, diffusers.models.attention_processor.IPAdapterXFormersAttnProcessor, diffusers.models.attention_processor.SD3IPAdapterJointAttnProcessor2_0, diffusers.models.attention_processor.PAGIdentitySelfAttnProcessor2_0, diffusers.models.attention_processor.PAGCFGIdentitySelfAttnProcessor2_0, diffusers.models.attention_processor.LoRAAttnProcessor, diffusers.models.attention_processor.LoRAAttnProcessor2_0, diffusers.models.attention_processor.LoRAXFormersAttnProcessor, diffusers.models.attention_processor.LoRAAttnAddedKVProcessor]]] )
设置用于计算注意力的注意力处理器。
Transformer2DModelOutput
class diffusers.models.modeling_outputs.Transformer2DModelOutput
< 来源 >( sample: torch.Tensor )
参数
- sample (形状为
(批大小, 通道数, 高度, 宽度)的torch.Tensor；如果 Transformer2DModel 是离散的，则为(批大小, 向量嵌入数 - 1, 潜在像素数)) — 在encoder_hidden_states输入上进行条件化的隐藏状态输出。如果是离散的，则返回未去噪潜在像素的概率分布。
Transformer2DModel 的输出。