Args: in_chans: Number of input image channels. num_classes: Number of classes for classification head. e

LCNet Essentially a MobileNet-V3 crossed with a MobileNet-V1 Paper: `PP-LCNet: A Lightweight CPU Convolutional Neural Network` - https://arx

Creates a TinyNet model.

With 50% prob, negate the value

Dataset factory method In parentheses after each arg are the type of dataset supported for each arg, one of: * folder - default, timm fold

Args: dataset: The image dataset to load. input_size: Target input size (channels, height, width) tuple or size scalar.

Prepare a batch for dynamic batching. Args: data (DataProto): The input data. max_token_len (int): The maximum token length

feature reduction (output stride) accessor

Select a subset of the DataProto via batch_keys and meta_info_keys Args: batch_keys (list, optional): a list of strings indicatin

Performs a single optimization step. Args: closure (callable, optional): A closure that reevaluates the model and return

Args: dim: Number of input channels. out_dim: Number of output channels (or 2 * dim if None) norm_layer:

Args: img_size (int, tuple): input image size in_chans (int): number of input channels patch_size (int):

Args: dim (int): Number of input channels. out_dim (int): Number of output channels (or 2 * dim if None)

Converts a token (str) in an id using the vocab.

Creates an EfficientNet-V2 base model Ref impl: https://github.com/google/automl/tree/master/efficientnetv2 Paper: `EfficientNetV2: Smaller

Creates a mnasnet-a1 model. Ref impl: https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet Paper: https://arxiv.org/pdf/1807

Creates a mnasnet-b1 model. Ref impl: https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet Paper: https://arxiv.org/pdf/1807

Append dict to a dict of list.

Average the policy loss. Args: values: `(torch.Tensor)` shape: (bs, n_rewards, response_length) mask: `(torch.Tensor)

Args: feat_shape: Spatial shape of the target tensor for embedding. bands: Optional pre-generated frequency bands dim: O

Fill a template with data.

Freeze parameters of the specified modules and those of all their hierarchical descendants. This is done in place. Args: root_module

Insert CLS token.