Initial attempt at implementing resnet50 for use with CIFAR data.

.gitignore

@@ -150,4 +150,7 @@ _html/
 .initialize_new_project.sh
 
 # Model files
-**/*.pth
+**/*.pth
+
+# Run results
+results/

.pre-commit-config.yaml

@@ -16,6 +16,7 @@ repos:
         name: Clear output from Jupyter notebooks
         description: Clear output from Jupyter notebooks.
         files: \.ipynb$
+        exclude: ^docs/pre_executed
         stages: [commit]
         language: system
         entry: jupyter nbconvert --clear-output

example_config.toml

@@ -17,6 +17,7 @@ log_level = "info"     # Emit informational messages, warnings and all errors
 # log_level = "debug"    # Very verbose, emit all log messages.
 
 data_dir = "/home/drew/code/fibad/data/"
+results_dir = "./results" # Results get named <verb>-<timestamp> under this directory
 
 [download]
 sw = "22asec"
@@ -52,20 +53,20 @@ mask = false
 [model]
 # The name of the built-in model to use or the libpath to an external model
 # e.g. "user_package.submodule.ExternalModel" or "ExampleAutoencoder"
-name = "kbmod_ml.models.cnn.CNN"
+name = "kbmod_ml.models.resnet50.RESNET50"
 
-weights_filepath = "example_model.pth"
+weights_filepath = "resnet50.pth"
 epochs = 10
 
-base_channel_size = 32
-latent_dim =64
+num_classes = 10
 
-[data_loader]
+
+[data_set]
 # Name of the built-in data loader to use or the libpath to an external data loader
 # e.g. "user_package.submodule.ExternalDataLoader" or "HSCDataLoader"
 name = "CifarDataLoader"
 
-
+[data_loader]
 # Pixel dimensions used to crop all images prior to loading. Will prune any images that are too small.
 #
 # If not provided by user, the default of 'false' scans the directory for the smallest dimensioned files, and 
@@ -83,9 +84,10 @@ crop_to = false
 filters = false
 
 # Default PyTorch DataLoader parameters
-batch_size = 4
+batch_size = 10
 shuffle = true
-num_workers = 2
+num_workers = 10
 
 [predict]
+model_weights_file = false
 batch_size = 32

src/kbmod_ml/models/resnet50.py

@@ -0,0 +1,65 @@
+# ruff: noqa: D101, D102
+
+# This example model is taken from the PyTorch CIFAR10 tutorial:
+# https://pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html#define-a-convolutional-neural-network
+import logging
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F  # noqa N812
+import torch.optim as optim
+from fibad.models.model_registry import fibad_model
+from torchvision.models import resnet50
+
+logger = logging.getLogger(__name__)
+
+
+@fibad_model
+class RESNET50(nn.Module):
+    def __init__(self, model_config, shape):
+        logger.info("This is an external model, not in FIBAD!!!")
+        super().__init__()
+
+        self.config = model_config
+
+        self.model = resnet50(pretrained=False, num_classes=self.config["model"]["num_classes"])
+
+        # Optimizer and criterion could be set directly, i.e. `self.optimizer = optim.SGD(...)`
+        # but we define them as methods as a way to allow for more flexibility in the future.
+        self.optimizer = self._optimizer()
+        self.criterion = self._criterion()
+
+    def forward(self, x):
+        return self.model(x)
+
+    def train_step(self, batch):
+        """This function contains the logic for a single training step. i.e. the
+        contents of the inner loop of a ML training process.
+
+        Parameters
+        ----------
+        batch : tuple
+            A tuple containing the inputs and labels for the current batch.
+
+        Returns
+        -------
+        Current loss value
+            The loss value for the current batch.
+        """
+        inputs, labels = batch
+
+        self.optimizer.zero_grad()
+        outputs = self(inputs)
+        loss = self.criterion(outputs, labels)
+        loss.backward()
+        self.optimizer.step()
+        return {"loss": loss.item()}
+
+    def _criterion(self):
+        return nn.CrossEntropyLoss()
+
+    def _optimizer(self):
+        return optim.SGD(self.parameters(), lr=0.001, momentum=0.9)
+
+    def save(self):
+        torch.save(self.state_dict(), self.config.get("weights_filepath"))