minerva.models.nets.image.setr ============================== .. py:module:: minerva.models.nets.image.setr Classes ------- .. autoapisummary:: minerva.models.nets.image.setr.ConvModule minerva.models.nets.image.setr.MMDropPath minerva.models.nets.image.setr.SETR_PUP minerva.models.nets.image.setr._SETRUPHead minerva.models.nets.image.setr._SetR_PUP Module Contents --------------- .. py:class:: ConvModule(in_channels, out_channels, kernel_size, padding, norm_type, act_type, norm_params = None, act_params = None) Bases: :py:obj:`torch.nn.Module` Base class for all neural network modules. Your models should also subclass this class. Modules can also contain other Modules, allowing them to be nested in a tree structure. You can assign the submodules as regular attributes:: import torch.nn as nn import torch.nn.functional as F class Model(nn.Module): def __init__(self) -> None: super().__init__() self.conv1 = nn.Conv2d(1, 20, 5) self.conv2 = nn.Conv2d(20, 20, 5) def forward(self, x): x = F.relu(self.conv1(x)) return F.relu(self.conv2(x)) Submodules assigned in this way will be registered, and will also have their parameters converted when you call :meth:`to`, etc. .. note:: As per the example above, an ``__init__()`` call to the parent class must be made before assignment on the child. :ivar training: Boolean represents whether this module is in training or evaluation mode. :vartype training: bool Convolutional module with normalization and activation. Parameters ---------- in_channels : int Number of input channels. out_channels : int Number of output channels. kernel_size : int Size of the convolution kernel. padding : int Padding added to both sides of the input. norm_type : type Type of normalization layer (e.g., nn.BatchNorm2d). act_type : type Type of activation function (e.g., nn.ReLU). norm_params : dict, optional Optional parameters for normalization. act_params : dict, optional Optional parameters for activation. .. py:attribute:: activate .. py:attribute:: bn .. py:attribute:: conv .. py:method:: forward(x) Forward pass of the ConvModule. .. py:method:: init_weights() Initialize convolution weights. .. py:class:: MMDropPath(drop_prob) Bases: :py:obj:`torch.nn.Module` Base class for all neural network modules. Your models should also subclass this class. Modules can also contain other Modules, allowing them to be nested in a tree structure. You can assign the submodules as regular attributes:: import torch.nn as nn import torch.nn.functional as F class Model(nn.Module): def __init__(self) -> None: super().__init__() self.conv1 = nn.Conv2d(1, 20, 5) self.conv2 = nn.Conv2d(20, 20, 5) def forward(self, x): x = F.relu(self.conv1(x)) return F.relu(self.conv2(x)) Submodules assigned in this way will be registered, and will also have their parameters converted when you call :meth:`to`, etc. .. note:: As per the example above, an ``__init__()`` call to the parent class must be made before assignment on the child. :ivar training: Boolean represents whether this module is in training or evaluation mode. :vartype training: bool Initialize internal Module state, shared by both nn.Module and ScriptModule. .. py:attribute:: drop_prob .. py:method:: forward(x) .. py:class:: SETR_PUP(original_resolution = None, img_size = (512, 512), patch_size = 16, in_channels = 3, embed_dims = 1024, num_layers = 24, num_heads = 16, out_indices = (9, 14, 19, 23), encoder_stride = None, patch_norm = False, dilatation = 1, bias = True, padding_type = 'corner', mlp_ratio = 4, attn_drop_rate = 0.0, drop_path_rate = 0.0, num_fcs = 2, qkv_bias = True, output_cls_token = False, act_type = nn.GELU, with_cp = False, encoder_dropout = 0.0, encoder_norm_type = nn.LayerNorm, dropout_type = MMDropPath, cls_token = True, interpolate_mode = 'bilinear', act_params = None, dropout_params = None, encoder_norm_params = None, decoder_channels = 256, decoder_in_index = 3, num_classes = 6, decoder_dropout = 0.0, decoder_norm_type = nn.SyncBatchNorm, decoder_num_convs = 4, decoder_up_scale = 2, decoder_kernel_size = 3, decoder_align_corners = False, decoder_norm_params = None, aux_heads_in_index = (0, 1, 2), aux_head_num_convs = 2, aux_head_up_scale = 4, aux_weights = None, loss_fn = None, optimizer = Adam, optimizer_kwargs = None, train_metrics = None, val_metrics = None, test_metrics = None, freeze_backbone = False, learning_rate = 0.001, loss_weights = None, lr_scheduler = None, lr_scheduler_kwargs = None, head_lr_factor = 1.0, use_sliding_inference = True, sliding_window_stride = (341, 341)) Bases: :py:obj:`minerva.models.nets.base.SimpleSupervisedModel` A modular Lightning model wrapper for supervised learning tasks. This class enables the construction of supervised models by combining a backbone (feature extractor), an optional adapter, and a fully connected (FC) head. It provides a clean interface for setting up custom training, validation, and testing pipelines with pluggable loss functions, metrics, optimizers, and learning rate schedulers. The architecture is structured as follows: +------------------+ | Backbone Model | +------------------+ | v +------------------------+ | Adapter (Optional) | +------------------------+ | (Flatten if needed) v +------------------------+ | Fully Connected Head | +------------------------+ | v +------------------+ | Loss Function | +------------------+ Training and validation steps comprise the following steps: 1. Forward pass input through the backbone. 2. Pass through adapter (if provided). 3. Flatten the output (if `flatten` is True) before the FC head. 4. Forward through the FC head. 5. Compute loss with respect to targets. 6. Backpropagate and update parameters. 7. Compute metrics and log them. 8. Return loss. `train_loss`, `val_loss`, and `test_loss` are always logged, along with any additional metrics specified in the `train_metrics`, `val_metrics`, and `test_metrics` dictionaries. This wrapper is especially useful to quickly set up supervised models for various tasks, such as image classification, object detection, and segmentation. It is designed to be flexible and extensible, allowing users to easily swap out components like the backbone, adapter, and FC head as needed. The model is built with a focus on simplicity and modularity, making it easy to adapt to different use cases and requirements. The model is designed to be used with PyTorch Lightning and is compatible with its training loop. **Note**: For more complex architectures that does not follow the above structure should not inherit from this class. **Note**: Input batches must be tuples (input_tensor, target_tensor). LightningModule implementation for SETR_PUP (SEgmentation TRansformer with Progressive UPsampling). Parameters ---------- original_resolution : tuple Resolution of the original images used to pretrain the backbone. img_size : tuple Input image size (height, width) used during training and patch embedding. patch_size : int Size of each image patch extracted in the ViT encoder. in_channels : int Number of input channels (usually 3 for RGB). embed_dims : int Embedding dimension for each patch. num_layers : int Number of transformer encoder layers. num_heads : int Number of attention heads in each transformer layer. out_indices : tuple, optional Indices of the encoder layers to use as features for decoding. encoder_stride : int Stride used in patch embedding. patch_norm : bool Whether to apply normalization to patch embeddings. dilatation : int Dilation factor for patch embedding. bias : bool Whether to include bias in the projection layers. padding_type : str Padding mode used in patch embedding ("same" or "corner"). mlp_ratio : int Expansion ratio for the MLP block inside transformer layers. attn_drop_rate : float Dropout rate applied to attention weights. drop_path_rate : float Probability of dropping entire residual paths (stochastic depth). num_fcs : int Number of linear layers in the feed-forward MLP of the transformer. qkv_bias : bool Whether to include bias in QKV projections. output_cls_token : bool Whether to include class token in encoder output. act_type : type Activation function class to use (e.g., nn.GELU). with_cp : bool Whether to enable checkpointing to save memory. encoder_dropout : float Dropout rate after positional embedding in the encoder. encoder_norm_type : type Normalization type used in the encoder. dropout_type : type Type of stochastic path dropout layer. cls_token : bool Whether to use a class token in the ViT. interpolate_mode : str Interpolation mode used for resizing positional embeddings. act_params : dict, optional Additional parameters for the activation function. dropout_params : dict, optional Additional parameters for the dropout layer. encoder_norm_params : dict, optional Additional parameters for the encoder normalization layer. decoder_channels : int Number of channels in intermediate layers of the decoder. decoder_in_index : int Index into encoder outputs to be used as decoder input. num_classes : int Number of segmentation classes. decoder_dropout : float Dropout probability in the decoder. decoder_norm_type : type Type of normalization in decoder conv blocks. decoder_num_convs : int Number of conv+upsample blocks in the decoder. decoder_up_scale : int Upsample scale factor for decoder blocks. decoder_kernel_size : int Convolution kernel size in decoder blocks. decoder_align_corners : bool Whether to align corners when using bilinear interpolation. decoder_norm_params : dict, optional Additional arguments for decoder normalization. aux_heads_in_index : tuple of int Indices of encoder layers to feed into each auxiliary decoder head. aux_head_num_convs : int Number of conv blocks in each auxiliary head. aux_head_up_scale : int Upsample factor for auxiliary heads. aux_weights : list of float, optional Weights for auxiliary losses [aux1, aux2, aux3]. loss_fn : nn.Module, optional Loss function module (defaults to CrossEntropy). optimizer_type : type, optional Optimizer class (e.g., torch.optim.Adam). optimizer_params : dict, optional Parameters to pass to the optimizer. train_metrics : dict, optional Dictionary of training metrics. val_metrics : dict, optional Dictionary of validation metrics. test_metrics : dict, optional Dictionary of test metrics. learning_rate : float Learning rate for training. loss_weights : list of float, optional Class-wise weights for the loss function. head_lr_factor : float Learning rate multiplier for decoder heads. lr_scheduler : type, optional Learning rate scheduler class to be instantiated. By default, it is set to None, which means no scheduler will be used. Should be a subclass of `torch.optim.lr_scheduler.LRScheduler` (e.g., `torch.optim.lr_scheduler.StepLR`). lr_scheduler_kwargs : dict, optional Additional kwargs passed to the scheduler constructor. use_sliding_inference : bool Whether to use sliding window inference for large images on validation and test. sliding_window_stride : tuple of int Stride for sliding window inference (height, width). .. py:method:: _compute_metrics(y_hat, y, step_name) Calculate the metrics for the given step. Parameters ---------- y_hat : torch.Tensor The output data from the forward pass. y : torch.Tensor The input data/label. step_name : str Name of the step. It will be used to get the metrics from the `self.metrics` attribute. Returns ------- Dict[str, torch.Tensor] A dictionary with the metrics values. .. py:method:: _eval_step_with_slide(batch, step_name) .. py:method:: _loss_func(y_hat, y) Calculate the loss between the output and the input data. Parameters ---------- y_hat : torch.Tensor The output data from the forward pass. y : torch.Tensor The input data/label. Returns ------- torch.Tensor The loss value. .. py:method:: _slide_inference(image, crop_size=(512, 512), stride=(341, 341), ori_shape = None) Realiza inferência por janelamento (sliding window) com reconstrução e resize final opcional. .. py:attribute:: aux_head1 .. py:attribute:: aux_head2 .. py:attribute:: aux_head3 .. py:attribute:: aux_weights :value: None .. py:method:: configure_optimizers() Choose what optimizers and learning-rate schedulers to use in your optimization. Normally you'd need one. But in the case of GANs or similar you might have multiple. Optimization with multiple optimizers only works in the manual optimization mode. Return: Any of these 6 options. - **Single optimizer**. - **List or Tuple** of optimizers. - **Two lists** - The first list has multiple optimizers, and the second has multiple LR schedulers (or multiple ``lr_scheduler_config``). - **Dictionary**, with an ``"optimizer"`` key, and (optionally) a ``"lr_scheduler"`` key whose value is a single LR scheduler or ``lr_scheduler_config``. - **None** - Fit will run without any optimizer. The ``lr_scheduler_config`` is a dictionary which contains the scheduler and its associated configuration. The default configuration is shown below. .. code-block:: python lr_scheduler_config = { # REQUIRED: The scheduler instance "scheduler": lr_scheduler, # The unit of the scheduler's step size, could also be 'step'. # 'epoch' updates the scheduler on epoch end whereas 'step' # updates it after a optimizer update. "interval": "epoch", # How many epochs/steps should pass between calls to # `scheduler.step()`. 1 corresponds to updating the learning # rate after every epoch/step. "frequency": 1, # Metric to monitor for schedulers like `ReduceLROnPlateau` "monitor": "val_loss", # If set to `True`, will enforce that the value specified 'monitor' # is available when the scheduler is updated, thus stopping # training if not found. If set to `False`, it will only produce a warning "strict": True, # If using the `LearningRateMonitor` callback to monitor the # learning rate progress, this keyword can be used to specify # a custom logged name "name": None, } When there are schedulers in which the ``.step()`` method is conditioned on a value, such as the :class:`torch.optim.lr_scheduler.ReduceLROnPlateau` scheduler, Lightning requires that the ``lr_scheduler_config`` contains the keyword ``"monitor"`` set to the metric name that the scheduler should be conditioned on. .. testcode:: # The ReduceLROnPlateau scheduler requires a monitor def configure_optimizers(self): optimizer = Adam(...) return { "optimizer": optimizer, "lr_scheduler": { "scheduler": ReduceLROnPlateau(optimizer, ...), "monitor": "metric_to_track", "frequency": "indicates how often the metric is updated", # If "monitor" references validation metrics, then "frequency" should be set to a # multiple of "trainer.check_val_every_n_epoch". }, } # In the case of two optimizers, only one using the ReduceLROnPlateau scheduler def configure_optimizers(self): optimizer1 = Adam(...) optimizer2 = SGD(...) scheduler1 = ReduceLROnPlateau(optimizer1, ...) scheduler2 = LambdaLR(optimizer2, ...) return ( { "optimizer": optimizer1, "lr_scheduler": { "scheduler": scheduler1, "monitor": "metric_to_track", }, }, {"optimizer": optimizer2, "lr_scheduler": scheduler2}, ) Metrics can be made available to monitor by simply logging it using ``self.log('metric_to_track', metric_val)`` in your :class:`~lightning.pytorch.core.LightningModule`. Note: Some things to know: - Lightning calls ``.backward()`` and ``.step()`` automatically in case of automatic optimization. - If a learning rate scheduler is specified in ``configure_optimizers()`` with key ``"interval"`` (default "epoch") in the scheduler configuration, Lightning will call the scheduler's ``.step()`` method automatically in case of automatic optimization. - If you use 16-bit precision (``precision=16``), Lightning will automatically handle the optimizer. - If you use :class:`torch.optim.LBFGS`, Lightning handles the closure function automatically for you. - If you use multiple optimizers, you will have to switch to 'manual optimization' mode and step them yourself. - If you need to control how often the optimizer steps, override the :meth:`optimizer_step` hook. .. py:method:: create_from_dict(config) :staticmethod: .. py:attribute:: decoder_num_classes :value: 6 .. py:attribute:: encoder_out_indices :value: (9, 14, 19, 23) .. py:method:: forward(x) Perform a forward pass with the input data on the backbone model. Parameters ---------- x : torch.Tensor The input data. Returns ------- torch.Tensor The output data from the forward pass. .. py:attribute:: head_lr_factor :value: 1.0 .. py:attribute:: img_size :value: (512, 512) .. py:method:: load_backbone(path, freeze = False) Loads pretrained ViT backbone optionally freezing its weights. .. py:attribute:: num_classes :value: 6 .. py:method:: predict_step(batch, batch_idx, dataloader_idx = None) Step function called during :meth:`~lightning.pytorch.trainer.trainer.Trainer.predict`. By default, it calls :meth:`~lightning.pytorch.core.LightningModule.forward`. Override to add any processing logic. The :meth:`~lightning.pytorch.core.LightningModule.predict_step` is used to scale inference on multi-devices. To prevent an OOM error, it is possible to use :class:`~lightning.pytorch.callbacks.BasePredictionWriter` callback to write the predictions to disk or database after each batch or on epoch end. The :class:`~lightning.pytorch.callbacks.BasePredictionWriter` should be used while using a spawn based accelerator. This happens for ``Trainer(strategy="ddp_spawn")`` or training on 8 TPU cores with ``Trainer(accelerator="tpu", devices=8)`` as predictions won't be returned. Args: batch: The output of your data iterable, normally a :class:`~torch.utils.data.DataLoader`. batch_idx: The index of this batch. dataloader_idx: The index of the dataloader that produced this batch. (only if multiple dataloaders used) Return: Predicted output (optional). Example :: class MyModel(LightningModule): def predict_step(self, batch, batch_idx, dataloader_idx=0): return self(batch) dm = ... model = MyModel() trainer = Trainer(accelerator="gpu", devices=2) predictions = trainer.predict(model, dm) .. py:method:: test_step(batch, batch_idx) Operates on a single batch of data from the test set. In this step you'd normally generate examples or calculate anything of interest such as accuracy. Args: batch: The output of your data iterable, normally a :class:`~torch.utils.data.DataLoader`. batch_idx: The index of this batch. dataloader_idx: The index of the dataloader that produced this batch. (only if multiple dataloaders used) Return: - :class:`~torch.Tensor` - The loss tensor - ``dict`` - A dictionary. Can include any keys, but must include the key ``'loss'``. - ``None`` - Skip to the next batch. .. code-block:: python # if you have one test dataloader: def test_step(self, batch, batch_idx): ... # if you have multiple test dataloaders: def test_step(self, batch, batch_idx, dataloader_idx=0): ... Examples:: # CASE 1: A single test dataset def test_step(self, batch, batch_idx): x, y = batch # implement your own out = self(x) loss = self.loss(out, y) # log 6 example images # or generated text... or whatever sample_imgs = x[:6] grid = torchvision.utils.make_grid(sample_imgs) self.logger.experiment.add_image('example_images', grid, 0) # calculate acc labels_hat = torch.argmax(out, dim=1) test_acc = torch.sum(y == labels_hat).item() / (len(y) * 1.0) # log the outputs! self.log_dict({'test_loss': loss, 'test_acc': test_acc}) If you pass in multiple test dataloaders, :meth:`test_step` will have an additional argument. We recommend setting the default value of 0 so that you can quickly switch between single and multiple dataloaders. .. code-block:: python # CASE 2: multiple test dataloaders def test_step(self, batch, batch_idx, dataloader_idx=0): # dataloader_idx tells you which dataset this is. x, y = batch # implement your own out = self(x) if dataloader_idx == 0: loss = self.loss0(out, y) else: loss = self.loss1(out, y) # calculate acc labels_hat = torch.argmax(out, dim=1) acc = torch.sum(y == labels_hat).item() / (len(y) * 1.0) # log the outputs separately for each dataloader self.log_dict({f"test_loss_{dataloader_idx}": loss, f"test_acc_{dataloader_idx}": acc}) Note: If you don't need to test you don't need to implement this method. Note: When the :meth:`test_step` is called, the model has been put in eval mode and PyTorch gradients have been disabled. At the end of the test epoch, the model goes back to training mode and gradients are enabled. .. py:method:: training_step(batch, batch_idx) Here you compute and return the training loss and some additional metrics for e.g. the progress bar or logger. Args: batch: The output of your data iterable, normally a :class:`~torch.utils.data.DataLoader`. batch_idx: The index of this batch. dataloader_idx: The index of the dataloader that produced this batch. (only if multiple dataloaders used) Return: - :class:`~torch.Tensor` - The loss tensor - ``dict`` - A dictionary which can include any keys, but must include the key ``'loss'`` in the case of automatic optimization. - ``None`` - In automatic optimization, this will skip to the next batch (but is not supported for multi-GPU, TPU, or DeepSpeed). For manual optimization, this has no special meaning, as returning the loss is not required. In this step you'd normally do the forward pass and calculate the loss for a batch. You can also do fancier things like multiple forward passes or something model specific. Example:: def training_step(self, batch, batch_idx): x, y, z = batch out = self.encoder(x) loss = self.loss(out, x) return loss To use multiple optimizers, you can switch to 'manual optimization' and control their stepping: .. code-block:: python def __init__(self): super().__init__() self.automatic_optimization = False # Multiple optimizers (e.g.: GANs) def training_step(self, batch, batch_idx): opt1, opt2 = self.optimizers() # do training_step with encoder ... opt1.step() # do training_step with decoder ... opt2.step() Note: When ``accumulate_grad_batches`` > 1, the loss returned here will be automatically normalized by ``accumulate_grad_batches`` internally. .. py:attribute:: use_sliding_inference :value: True .. py:method:: validation_step(batch, batch_idx) Operates on a single batch of data from the validation set. In this step you'd might generate examples or calculate anything of interest like accuracy. Args: batch: The output of your data iterable, normally a :class:`~torch.utils.data.DataLoader`. batch_idx: The index of this batch. dataloader_idx: The index of the dataloader that produced this batch. (only if multiple dataloaders used) Return: - :class:`~torch.Tensor` - The loss tensor - ``dict`` - A dictionary. Can include any keys, but must include the key ``'loss'``. - ``None`` - Skip to the next batch. .. code-block:: python # if you have one val dataloader: def validation_step(self, batch, batch_idx): ... # if you have multiple val dataloaders: def validation_step(self, batch, batch_idx, dataloader_idx=0): ... Examples:: # CASE 1: A single validation dataset def validation_step(self, batch, batch_idx): x, y = batch # implement your own out = self(x) loss = self.loss(out, y) # log 6 example images # or generated text... or whatever sample_imgs = x[:6] grid = torchvision.utils.make_grid(sample_imgs) self.logger.experiment.add_image('example_images', grid, 0) # calculate acc labels_hat = torch.argmax(out, dim=1) val_acc = torch.sum(y == labels_hat).item() / (len(y) * 1.0) # log the outputs! self.log_dict({'val_loss': loss, 'val_acc': val_acc}) If you pass in multiple val dataloaders, :meth:`validation_step` will have an additional argument. We recommend setting the default value of 0 so that you can quickly switch between single and multiple dataloaders. .. code-block:: python # CASE 2: multiple validation dataloaders def validation_step(self, batch, batch_idx, dataloader_idx=0): # dataloader_idx tells you which dataset this is. x, y = batch # implement your own out = self(x) if dataloader_idx == 0: loss = self.loss0(out, y) else: loss = self.loss1(out, y) # calculate acc labels_hat = torch.argmax(out, dim=1) acc = torch.sum(y == labels_hat).item() / (len(y) * 1.0) # log the outputs separately for each dataloader self.log_dict({f"val_loss_{dataloader_idx}": loss, f"val_acc_{dataloader_idx}": acc}) Note: If you don't need to validate you don't need to implement this method. Note: When the :meth:`validation_step` is called, the model has been put in eval mode and PyTorch gradients have been disabled. At the end of validation, the model goes back to training mode and gradients are enabled. .. py:class:: _SETRUPHead(in_channels, channels, num_classes, in_index, num_convs, up_scale, kernel_size, align_corners, dropout, norm_type, act_type, norm_params = None, act_params = None, interpolate_mode = 'bilinear') Bases: :py:obj:`torch.nn.Module` Base class for all neural network modules. Your models should also subclass this class. Modules can also contain other Modules, allowing them to be nested in a tree structure. You can assign the submodules as regular attributes:: import torch.nn as nn import torch.nn.functional as F class Model(nn.Module): def __init__(self) -> None: super().__init__() self.conv1 = nn.Conv2d(1, 20, 5) self.conv2 = nn.Conv2d(20, 20, 5) def forward(self, x): x = F.relu(self.conv1(x)) return F.relu(self.conv2(x)) Submodules assigned in this way will be registered, and will also have their parameters converted when you call :meth:`to`, etc. .. note:: As per the example above, an ``__init__()`` call to the parent class must be made before assignment on the child. :ivar training: Boolean represents whether this module is in training or evaluation mode. :vartype training: bool Lightweight decoder head with LayerNorm and upsampling for SETR. Parameters ---------- in_channels : int Number of input channels from encoder. channels : int Number of internal intermediate channels. num_classes : int Number of target output classes. in_index : int Index to select feature from encoder outputs. num_convs : int Number of upsampling convolutional layers. up_scale : int Upsample factor per layer. kernel_size : int Convolution kernel size. align_corners : bool Align corners in bilinear upsampling. dropout : float Dropout probability. norm_type : type Normalization layer type. act_type : type Activation function type. norm_params : dict, optional Additional parameters for normalization. act_params : dict, optional Additional parameters for activation. interpolate_mode : str, default="bilinear" Interpolation mode for upsampling. .. py:attribute:: align_corners .. py:attribute:: conv_seg .. py:method:: forward(xs) Forward pass of SETR decoder head. .. py:attribute:: in_index .. py:attribute:: norm .. py:attribute:: up_convs .. py:class:: _SetR_PUP(original_resolution, img_size, patch_size, in_channels, embed_dims, num_layers, num_heads, out_indices, stride, patch_norm, dilatation, bias, padding_type, mlp_ratio, attn_drop_rate, drop_path_rate, num_fcs, qkv_bias, output_cls_token, act_type, with_cp, encoder_dropout, encoder_norm_type, dropout_type, cls_token, interpolate_mode, act_params, dropout_params, encoder_norm_params, decoder_channels, decoder_in_index, num_classes, decoder_dropout, decoder_norm_type, decoder_num_convs, decoder_up_scale, decoder_kernel_size, decoder_align_corners, decoder_norm_params, aux_heads_in_index, aux_head_num_convs, aux_head_up_scale) Bases: :py:obj:`torch.nn.Module` Base class for all neural network modules. Your models should also subclass this class. Modules can also contain other Modules, allowing them to be nested in a tree structure. You can assign the submodules as regular attributes:: import torch.nn as nn import torch.nn.functional as F class Model(nn.Module): def __init__(self) -> None: super().__init__() self.conv1 = nn.Conv2d(1, 20, 5) self.conv2 = nn.Conv2d(20, 20, 5) def forward(self, x): x = F.relu(self.conv1(x)) return F.relu(self.conv2(x)) Submodules assigned in this way will be registered, and will also have their parameters converted when you call :meth:`to`, etc. .. note:: As per the example above, an ``__init__()`` call to the parent class must be made before assignment on the child. :ivar training: Boolean represents whether this module is in training or evaluation mode. :vartype training: bool Full SETR_PUP model with encoder and decoder. Parameters ---------- original_resolution : tuple Resolution of original input image. img_size : tuple Input image size used for patch embedding. patch_size : int Patch size for Vision Transformer. in_channels : int Number of input image channels. embed_dims : int Dimensionality of embeddings. num_layers : int Number of transformer encoder layers. num_heads : int Number of attention heads. out_indices : tuple or None Indices of intermediate outputs for decoding. stride : int Patch stride. patch_norm : bool Apply normalization to patches. dilatation : int Dilation for patch embedding. bias : bool Use bias in conv layers. padding_type : str Padding type used for patch embedding. mlp_ratio : int MLP expansion ratio. attn_drop_rate : float Attention dropout rate. drop_path_rate : float Stochastic depth dropout rate. num_fcs : int Number of fully connected layers in FFN. qkv_bias : bool Use bias in QKV projections. output_cls_token : bool Output class token with final features. act_type : type Activation function type. with_cp : bool Use gradient checkpointing. encoder_dropout : float Dropout rate after patch embedding. encoder_norm_type : type Type of normalization used in encoder. dropout_type : type Type of residual dropout layer. cls_token : bool Use class token in transformer. interpolate_mode : str Mode for interpolating positional embeddings. act_params : dict, optional Params for activation function. dropout_params : dict, optional Params for dropout module. encoder_norm_params : dict, optional Params for encoder normalization. decoder_channels : int Number of intermediate decoder channels. decoder_in_index : int Which encoder layer to use in decoder. num_classes : int Number of classes for segmentation. decoder_dropout : float Dropout rate in decoder. decoder_norm_type : type Normalization type in decoder. decoder_num_convs : int Number of conv blocks in decoder. decoder_up_scale : int Upsample scale factor. decoder_kernel_size : int Decoder conv kernel size. decoder_align_corners : bool Use align_corners in bilinear upsample. decoder_norm_params : dict, optional Parameters for decoder normalization. aux_heads_in_index : tuple of int Which layers to use in auxiliary decoders. aux_head_num_convs : int Number of convs in each auxiliary head. aux_head_up_scale : int Upsample factor in each auxiliary head. .. py:attribute:: aux_head1 .. py:attribute:: aux_head2 .. py:attribute:: aux_head3 .. py:attribute:: decoder .. py:attribute:: encoder .. py:attribute:: encoder_out_indices .. py:method:: forward(x)