How txstcodeplayground/codecs4337fall2026 works
This repository is a collection of self-contained CS4337 teaching projects for training and inspecting vision models, not one unified application. Its main workflow is to select a project and config, load a dataset, construct and train or evaluate a model, then inspect local results or tracking dashboards. The graph keeps the documented MLP/ConvNet progression, hyperparameter-search projects, and the U-Net, transfer-learning, and YOLO examples distinct. Edges reflect README-documented capabilities and progression; unsampled implementations are not inferred in detail.
Main entry points: HyperparameterSearchConvnets/main.py ConvNet Search Run · LitWBTransferLearningResnet/main.py Flowers-102 Transfer · UnetSemanticSegmentationExample/main.py U-Net Segmentation · TensorBoardTrainingBasicNeuralNetwork/main.py TensorBoard MLP
Components 34
Training Foundations
- Plain MLP
TrainingBasicNeuralNetwork/main.py - MLP Model
TrainingBasicNeuralNetwork/models/mlp.py - MLP Trainer
TrainingBasicNeuralNetwork/trainers/trainer.py - Plain ConvNet
TrainingBasicConvnet/main.py - ConvNet Trainer
TrainingBasicConvnet/trainers/trainer.py
Lightning Training
- TensorBoard MLP
TensorBoardTrainingBasicNeuralNetwork/main.py - TensorBoard Trainer
TensorBoardTrainingBasicNeuralNetwork/trainers/trainer.py - Lightning MLP
LitTrainingBasicNeuralNetwork/main.py - Lightning ConvNet
LitTrainingBasicConvnet/main.py - W&B MLP
LitWBTrainingBasicNeuralNetwork/main.py - W&B ConvNet
LitWBTrainingBasicConvnet/main.py
Hyperparameter Search
- MLP Search
LitWBHSTrainingBasicNeuralNetwork/search.py - MLP Search Run
LitWBHSTrainingBasicNeuralNetwork/main.py - ConvNet Search
HyperparameterSearchConvnets/search.py - ConvNet Search Run
HyperparameterSearchConvnets/main.py - ConvNet Blocks
HyperparameterSearchConvnets/models/components/convnet.py - Search Lightning Model
HyperparameterSearchConvnets/models/lit_convnet.py - Search Data Loader
HyperparameterSearchConvnets/dataloaders/fashion_mnist.py
Vision Applications
- Flowers-102 Transfer
LitWBTransferLearningResnet/main.py - Flowers Data Loader
LitWBTransferLearningResnet/dataloaders/flowers102.py - ResNet Model
LitWBTransferLearningResnet/models/components/resnet.py - Evaluation Reports
LitWBTransferLearningResnet/callbacks/wandb_evaluation.py - U-Net Segmentation
UnetSemanticSegmentationExample/main.py - Kvasir Data Loader
UnetSemanticSegmentationExample/dataloaders/kvasir_seg.py - U-Net Models
UnetSemanticSegmentationExample/models/components/unet.py - Segmentation Reports
UnetSemanticSegmentationExample/callbacks/wandb_segmentation.py - YOLO Detection
YoloExample/main.py - Penn-Fudan Data
YoloExample/dataloaders/pennfudan.py - YOLO Model Setup
YoloExample/models/loading.py
Experiment Outputs
- TensorBoard Logging
TensorBoardTrainingBasicNeuralNetwork/utils/tensorboard_logging.py
Other components
- Learner
- Fashion-MNIST
- Weights & Biases
- TensorBoard Dashboard
Connections 43
- Learner → Plain MLP: runs project
README.md - Learner → Plain ConvNet: runs project
README.md - Learner → TensorBoard MLP: runs project
README.md - Learner → Lightning MLP: runs project
README.md - Learner → Lightning ConvNet: runs project
README.md - Learner → W&B MLP: runs project
README.md - Learner → W&B ConvNet: runs project
README.md - Learner → MLP Search Run: runs search
README.md - Learner → ConvNet Search Run: runs search
README.md - Learner → Flowers-102 Transfer: runs project
README.md - Learner → U-Net Segmentation: runs project
README.md - Learner → YOLO Detection: runs project
README.md - Fashion-MNIST → Plain MLP: training data
README.md - Fashion-MNIST → Plain ConvNet: training data
README.md - MLP Trainer → MLP Model: trains model
- Plain ConvNet → ConvNet Trainer: uses trainer
TrainingBasicConvnet/main.py - Plain ConvNet → Fashion-MNIST: trains on
README.md - TensorBoard MLP → TensorBoard Trainer: trains with
TensorBoardTrainingBasicNeuralNetwork/main.py - TensorBoard MLP → TensorBoard Logging: logs metrics
TensorBoardTrainingBasicNeuralNetwork/main.py - TensorBoard Logging → TensorBoard Dashboard: view events
- TensorBoard MLP → W&B MLP: builds on approach
README.md - Lightning MLP → W&B MLP: adds tracking
README.md - Lightning ConvNet → W&B ConvNet: adds tracking
README.md - W&B MLP → Weights & Biases: tracks run
README.md - W&B ConvNet → Weights & Biases: tracks run
README.md - MLP Search → MLP Search Run: searches parameters
README.md - MLP Search Run → Weights & Biases: logs best trial
README.md - ConvNet Search → ConvNet Search Run: searches architecture
README.md - ConvNet Search Run → Search Data Loader: loads batches
- ConvNet Search Run → ConvNet Blocks: builds network
- ConvNet Search Run → Search Lightning Model: trains model
- ConvNet Search Run → Weights & Biases: logs results
HyperparameterSearchConvnets/main.py - Flowers-102 Transfer → Flowers Data Loader: loads dataset
- Flowers-102 Transfer → ResNet Model: trains network
- Flowers-102 Transfer → Evaluation Reports: reports metrics
LitWBTransferLearningResnet/main.py - Flowers-102 Transfer → Weights & Biases: logs report
README.md - U-Net Segmentation → Kvasir Data Loader: loads dataset
- U-Net Segmentation → U-Net Models: trains model
- U-Net Segmentation → Segmentation Reports: reports predictions
README.md - U-Net Segmentation → Weights & Biases: tracks experiment
README.md - YOLO Detection → Penn-Fudan Data: prepares data
YoloExample/main.py - YOLO Detection → YOLO Model Setup: loads weights
YoloExample/main.py - YOLO Detection → Weights & Biases: tracks run
README.md