Buckets:
| # Fungsi aktivasi | |
| Fungsi aktivasi yang dikustomisasi untuk mendukung berbagai model di 🤗 Diffusers. | |
| ## GELU[[diffusers.models.activations.GELU]] | |
| #### diffusers.models.activations.GELU[[diffusers.models.activations.GELU]] | |
| ```python | |
| diffusers.models.activations.GELU(dim_in: int, dim_out: int, approximate: str = 'none', bias: bool = True) | |
| ``` | |
| [Sumber](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/activations.py#L65) | |
| **Parameter:** | |
| dim_in (`int`) : Jumlah channel pada input. | |
| dim_out (`int`) : Jumlah channel pada output. | |
| approximate (`str`, *opsional*, defaultnya `"none"`) : Jika `"tanh"`, gunakan aproksimasi tanh. | |
| bias (`bool`, defaultnya True) : Apakah akan menggunakan bias pada lapisan linear. | |
| Fungsi aktivasi GELU dengan dukungan aproksimasi tanh menggunakan `approximate="tanh"`. | |
| ## GEGLU[[diffusers.models.activations.GEGLU]] | |
| #### diffusers.models.activations.GEGLU[[diffusers.models.activations.GEGLU]] | |
| ```python | |
| diffusers.models.activations.GEGLU(dim_in: int, dim_out: int, bias: bool = True) | |
| ``` | |
| [Sumber](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/activations.py#L90) | |
| **Parameter:** | |
| dim_in (`int`) : Jumlah channel pada input. | |
| dim_out (`int`) : Jumlah channel pada output. | |
| bias (`bool`, defaultnya True) : Apakah akan menggunakan bias pada lapisan linear. | |
| Sebuah [varian](https://huggingface.co/papers/2002.05202) dari fungsi aktivasi gated linear unit. | |
| ## ApproximateGELU[[diffusers.models.activations.ApproximateGELU]] | |
| #### diffusers.models.activations.ApproximateGELU[[diffusers.models.activations.ApproximateGELU]] | |
| ```python | |
| diffusers.models.activations.ApproximateGELU(dim_in: int, dim_out: int, bias: bool = True) | |
| ``` | |
| [Sumber](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/activations.py#L143) | |
| **Parameter:** | |
| dim_in (`int`) : Jumlah channel pada input. | |
| dim_out (`int`) : Jumlah channel pada output. | |
| bias (`bool`, defaultnya True) : Apakah akan menggunakan bias pada lapisan linear. | |
| Bentuk aproksimasi dari Gaussian Error Linear Unit (GELU). Untuk detail lebih lanjut, lihat bagian 2 dari [makalah](https://huggingface.co/papers/1606.08415) ini. | |
| ## SwiGLU[[diffusers.models.activations.SwiGLU]] | |
| #### diffusers.models.activations.SwiGLU[[diffusers.models.activations.SwiGLU]] | |
| ```python | |
| diffusers.models.activations.SwiGLU(dim_in: int, dim_out: int, bias: bool = True) | |
| ``` | |
| [Sumber](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/activations.py#L120) | |
| **Parameter:** | |
| dim_in (`int`) : Jumlah channel pada input. | |
| dim_out (`int`) : Jumlah channel pada output. | |
| bias (`bool`, defaultnya True) : Apakah akan menggunakan bias pada lapisan linear. | |
| Sebuah [varian](https://huggingface.co/papers/2002.05202) dari fungsi aktivasi gated linear unit. Fungsinya mirip dengan `GEGLU` tetapi menggunakan SiLU / Swish alih-alih GeLU. | |
| ## FP32SiLU[[diffusers.models.activations.FP32SiLU]] | |
| #### diffusers.models.activations.FP32SiLU[[diffusers.models.activations.FP32SiLU]] | |
| ```python | |
| diffusers.models.activations.FP32SiLU() | |
| ``` | |
| [Sumber](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/activations.py#L53) | |
| Fungsi aktivasi SiLU dengan input yang di-upcast ke torch.float32. | |
| ## LinearActivation[[diffusers.models.activations.LinearActivation]] | |
| #### diffusers.models.activations.LinearActivation[[diffusers.models.activations.LinearActivation]] | |
| ```python | |
| diffusers.models.activations.LinearActivation(dim_in: int, dim_out: int, bias: bool = True, activation: str = 'silu') | |
| ``` | |
| [Sumber](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/activations.py#L163) | |
Xet Storage Details
- Size:
- 3.65 kB
- Xet hash:
- 0a3e7f33b62785b0525987c07a9a8dc9655ec96791e738741e593888b15b8a74
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.