fiftyone.utils.clip.model.VisionTransformer

class documentation

class VisionTransformer(nn.Module): (source)

Constructor: VisionTransformer(input_resolution, patch_size, width, layers, ...)

View In Hierarchy

Undocumented

Method __init__ Undocumented
Method forward Undocumented
Instance Variable class_embedding Undocumented
Instance Variable conv1 Undocumented
Instance Variable input_resolution Undocumented
Instance Variable ln_post Undocumented
Instance Variable ln_pre Undocumented
Instance Variable output_dim Undocumented
Instance Variable positional_embedding Undocumented
Instance Variable proj Undocumented
Instance Variable transformer Undocumented
def __init__(self, input_resolution: int, patch_size: int, width: int, layers: int, heads: int, output_dim: int): (source)

Undocumented

def forward(self, x: torch.Tensor): (source)

Undocumented

class_embedding = (source)

Undocumented

conv1 = (source)

Undocumented

input_resolution = (source)

Undocumented

ln_post = (source)

Undocumented

ln_pre = (source)

Undocumented

output_dim = (source)

Undocumented

positional_embedding = (source)

Undocumented

proj = (source)

Undocumented

transformer = (source)

Undocumented