# ErnieImageTransformer2DModel

用于 [ERNIE-Image](https://huggingface.co/baidu/ERNIE-Image) 中类图像数据的 Transformer 模型。

用于类图像数据的 Transformer 模型，来自 [ERNIE-Image-Turbo](https://huggingface.co/baidu/ERNIE-Image-Turbo)。

## ErnieImageTransformer2DModel[[diffusers.ErnieImageTransformer2DModel]]

#### diffusers.ErnieImageTransformer2DModel[[diffusers.ErnieImageTransformer2DModel]]

```python
diffusers.ErnieImageTransformer2DModel(hidden_size: int = 3072, num_attention_heads: int = 24, num_layers: int = 24, ffn_hidden_size: int = 8192, in_channels: int = 128, out_channels: int = 128, patch_size: int = 1, text_in_dim: int = 2560, rope_theta: int = 256, rope_axes_dim: typing.Tuple[int, int, int] = (32, 48, 48), eps: float = 1e-06, qk_layernorm: bool = True)
```

[源代码](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/transformers/transformer_ernie_image.py#L296)

#### forward[[diffusers.ErnieImageTransformer2DModel.forward]]

```python
forward(hidden_states: Tensor, timestep: Tensor, text_bth: Tensor, text_lens: Tensor, return_dict: bool = True)
```

[源代码](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/transformers/transformer_ernie_image.py#L348)

**参数：**

hidden_states (`torch.Tensor`，形状为 `(batch_size, in_channels, height, width)`)：输入 `hidden_states`。

timestep (`torch.LongTensor`)：用于指示去噪步骤。

text_bth (`torch.Tensor`) : 要使用的条件文本嵌入（从输入条件（例如提示词）计算得出的嵌入），形状为 `(batch_size, text_length, embed_dims)`。

text_lens (`torch.Tensor`) : 用于构建注意力掩码的每个样本的文本序列长度。

return_dict (`bool`，*可选*，默认为 `True`)：是否返回 `~models.transformer_2d.Transformer2DModelOutput` 而不是普通元组。

**返回值：**

如果 `return_dict` 为 True，则返回 `ErnieImageTransformer2DModelOutput`，否则返回一个 `tuple`，其中第一个元素是样本张量。

[ErnieImageTransformer2DModel](/docs/diffusers/main/zh/api/models/ernie_image_transformer2d#diffusers.ErnieImageTransformer2DModel) 的前向方法。