Buckets:
| # 归一化层 | |
| 用于支持 🤗 Diffusers 中各种模型的自定义归一化层。 | |
| ## AdaLayerNorm[[diffusers.models.normalization.AdaLayerNorm]] | |
| #### diffusers.models.normalization.AdaLayerNorm[[diffusers.models.normalization.AdaLayerNorm]] | |
| ```python | |
| diffusers.models.normalization.AdaLayerNorm(embedding_dim: int, num_embeddings: int | None = None, output_dim: int | None = None, norm_elementwise_affine: bool = False, norm_eps: float = 1e-05, chunk_dim: int = 0) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L27) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| num_embeddings (`int`, *可选*) : 嵌入字典的大小。 | |
| output_dim (`int`，*可选*) -- | |
| norm_elementwise_affine (`bool`，默认为 `False) -- | |
| norm_eps (`bool`，默认为 `False`) -- | |
| chunk_dim (`int`，默认为 `0`) -- | |
| 修改了 Norm 层以纳入时间步嵌入。 | |
| ## AdaLayerNormZero[[diffusers.models.normalization.AdaLayerNormZero]] | |
| #### diffusers.models.normalization.AdaLayerNormZero[[diffusers.models.normalization.AdaLayerNormZero]] | |
| ```python | |
| diffusers.models.normalization.AdaLayerNormZero(embedding_dim: int, num_embeddings: int | None = None, norm_type = 'layer_norm', bias = True) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L130) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| num_embeddings (`int`) : 嵌入字典的大小。 | |
| Norm 层自适应层归一化零（adaLN-Zero）。 | |
| ## AdaLayerNormSingle[[diffusers.models.normalization.AdaLayerNormSingle]] | |
| #### diffusers.models.normalization.AdaLayerNormSingle[[diffusers.models.normalization.AdaLayerNormSingle]] | |
| ```python | |
| diffusers.models.normalization.AdaLayerNormSingle(embedding_dim: int, use_additional_conditions: bool = False) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L235) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| use_additional_conditions (`bool`) : 是否使用额外的条件进行归一化。 | |
| 归一化层 自适应层归一化单（adaLN-single）。 | |
| 如 PixArt-Alpha 中所提出（参见：https://huggingface.co/papers/2310.00426；第 2.3 节）。 | |
| ## AdaGroupNorm[[diffusers.models.normalization.AdaGroupNorm]] | |
| #### diffusers.models.normalization.AdaGroupNorm[[diffusers.models.normalization.AdaGroupNorm]] | |
| ```python | |
| diffusers.models.normalization.AdaGroupNorm(embedding_dim: int, out_dim: int, num_groups: int, act_fn: str | None = None, eps: float = 1e-05) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L269) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| num_embeddings (`int`) : 嵌入字典的大小。 | |
| num_groups (`int`) : 将通道分组为多少个组。 | |
| act_fn (`str`，*可选*，默认为 `None`)：要使用的激活函数。 | |
| eps (`float`，*可选*，默认为 `1e-5`)：用于数值稳定性的 epsilon 值。 | |
| 修改后的 GroupNorm 层，用于整合时间步嵌入。 | |
| ## AdaLayerNormContinuous[[diffusers.models.normalization.AdaLayerNormContinuous]] | |
| #### diffusers.models.normalization.AdaLayerNormContinuous[[diffusers.models.normalization.AdaLayerNormContinuous]] | |
| ```python | |
| diffusers.models.normalization.AdaLayerNormContinuous(embedding_dim: int, conditioning_embedding_dim: int, elementwise_affine = True, eps = 1e-05, bias = True, norm_type = 'layer_norm') | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L307) | |
| **参数：** | |
| embedding_dim (`int`) : 投影过程中使用的嵌入维度。 | |
| conditioning_embedding_dim (`int`) : 输入条件的维度。 | |
| elementwise_affine (`bool`，默认为 `True`)：布尔标志，用于指示是否应应用仿射变换。 | |
| eps (`float`，默认值为 1e-5)：Epsilon 因子。 | |
| bias (`bias`，默认为 `True`)：布尔标志，用于指示是否使用偏置。 | |
| norm_type (`str`，默认为 `"layer_norm"`)：要使用的归一化层。支持的取值："layer_norm"、"rms_norm"。 | |
| 带有归一化层（layer_norm 或 rms_norm）的自适应归一化层。 | |
| ## RMSNorm[[diffusers.models.normalization.RMSNorm]] | |
| #### diffusers.models.normalization.RMSNorm[[diffusers.models.normalization.RMSNorm]] | |
| ```python | |
| diffusers.models.normalization.RMSNorm(dim, eps: float, elementwise_affine: bool = True, bias: bool = False) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L472) | |
| **参数：** | |
| dim (`int`) : 用于 `weights` 的维度数量。仅在 `elementwise_affine` 为 True 时生效。 | |
| eps (`float`) : 在计算平方根的倒数时使用的较小值。 | |
| elementwise_affine (`bool`，默认为 `True`)：布尔标志，用于指示是否应应用仿射变换。 | |
| bias (`bool`，默认为 False)：是否同时训练 `bias` 参数。 | |
| Zhang 等人于 https://huggingface.co/papers/1910.07467 中提出的 RMS Norm | |
| ## GlobalResponseNorm[[diffusers.models.normalization.GlobalResponseNorm]] | |
| #### diffusers.models.normalization.GlobalResponseNorm[[diffusers.models.normalization.GlobalResponseNorm]] | |
| ```python | |
| diffusers.models.normalization.GlobalResponseNorm(dim) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L563) | |
| **参数：** | |
| dim (`int`) : 用于 `gamma` 和 `beta` 的维度数量。 | |
| 在 ConvNeXt-v2 (https://huggingface.co/papers/2301.00808) 中引入的全局响应归一化。 | |
| ## LuminaLayerNormContinuous[[diffusers.models.normalization.LuminaLayerNormContinuous]] | |
| #### diffusers.models.normalization.LuminaLayerNormContinuous[[diffusers.models.normalization.LuminaLayerNormContinuous]] | |
| ```python | |
| diffusers.models.normalization.LuminaLayerNormContinuous(embedding_dim: int, conditioning_embedding_dim: int, elementwise_affine = True, eps = 1e-05, bias = True, norm_type = 'layer_norm', out_dim: int | None = None) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L354) | |
| ## SD35AdaLayerNormZeroX[[diffusers.models.normalization.SD35AdaLayerNormZeroX]] | |
| #### diffusers.models.normalization.SD35AdaLayerNormZeroX[[diffusers.models.normalization.SD35AdaLayerNormZeroX]] | |
| ```python | |
| diffusers.models.normalization.SD35AdaLayerNormZeroX(embedding_dim: int, norm_type: str = 'layer_norm', bias: bool = True) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L96) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| num_embeddings (`int`) : 嵌入字典的大小。 | |
| 归一化层自适应层归一化零（AdaLN-Zero）。 | |
| ## AdaLayerNormZeroSingle[[diffusers.models.normalization.AdaLayerNormZeroSingle]] | |
| #### diffusers.models.normalization.AdaLayerNormZeroSingle[[diffusers.models.normalization.AdaLayerNormZeroSingle]] | |
| ```python | |
| diffusers.models.normalization.AdaLayerNormZeroSingle(embedding_dim: int, norm_type = 'layer_norm', bias = True) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L173) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| num_embeddings (`int`) : 嵌入字典的大小。 | |
| Norm 层自适应层归一化零（adaLN-Zero）。 | |
| ## LuminaRMSNormZero[[diffusers.models.normalization.LuminaRMSNormZero]] | |
| #### diffusers.models.normalization.LuminaRMSNormZero[[diffusers.models.normalization.LuminaRMSNormZero]] | |
| ```python | |
| diffusers.models.normalization.LuminaRMSNormZero(embedding_dim: int, norm_eps: float, norm_elementwise_affine: bool) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L205) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| Norm 层自适应 RMS 归一化零。 | |
| ## LpNorm[[diffusers.models.normalization.LpNorm]] | |
| #### diffusers.models.normalization.LpNorm[[diffusers.models.normalization.LpNorm]] | |
| ```python | |
| diffusers.models.normalization.LpNorm(p: int = 2, dim: int = -1, eps: float = 1e-12) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L583) | |
| ## CogView3PlusAdaLayerNormZeroTextImage[[diffusers.models.normalization.CogView3PlusAdaLayerNormZeroTextImage]] | |
| #### diffusers.models.normalization.CogView3PlusAdaLayerNormZeroTextImage[[diffusers.models.normalization.CogView3PlusAdaLayerNormZeroTextImage]] | |
| ```python | |
| diffusers.models.normalization.CogView3PlusAdaLayerNormZeroTextImage(embedding_dim: int, dim: int) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L403) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| num_embeddings (`int`) : 嵌入字典的大小。 | |
| Norm 层自适应层归一化零（adaLN-Zero）。 | |
| ## CogVideoXLayerNormZero[[diffusers.models.normalization.CogVideoXLayerNormZero]] | |
| #### diffusers.models.normalization.CogVideoXLayerNormZero[[diffusers.models.normalization.CogVideoXLayerNormZero]] | |
| ```python | |
| diffusers.models.normalization.CogVideoXLayerNormZero(conditioning_dim: int, embedding_dim: int, elementwise_affine: bool = True, eps: float = 1e-05, bias: bool = True) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L448) | |
| ## MochiRMSNormZero[[diffusers.models.transformers.transformer_mochi.MochiRMSNormZero]] | |
| #### diffusers.models.transformers.transformer_mochi.MochiRMSNormZero[[diffusers.models.transformers.transformer_mochi.MochiRMSNormZero]] | |
| ```python | |
| diffusers.models.transformers.transformer_mochi.MochiRMSNormZero(embedding_dim: int, hidden_dim: int, eps: float = 1e-05, elementwise_affine: bool = False) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/transformers/transformer_mochi.py#L88) | |
| **参数：** | |
| embedding_dim (`int`) : 每个嵌入向量的维度。 | |
| Mochi 中使用的自适应 RMS Norm。 | |
| ## MochiRMSNorm[[diffusers.models.normalization.MochiRMSNorm]] | |
| #### diffusers.models.normalization.MochiRMSNorm[[diffusers.models.normalization.MochiRMSNorm]] | |
| ```python | |
| diffusers.models.normalization.MochiRMSNorm(dim, eps: float, elementwise_affine: bool = True) | |
| ``` | |
| [源代码](https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/models/normalization.py#L535) | |
Xet Storage Details
- Size:
- 10.4 kB
- Xet hash:
- b221ac4e8900f136d3f273fe1c27c7f1e04770a8868ee3ad0c1a42c9727b2fab
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.