tensorflow deep learning
via PatrickJS/awesome-cursorrules
TensorFlow best practices for building, training, and deploying neural networks at scale.
What is tensorflow deep learning?
Comprehensive rules for TensorFlow and deep learning projects covering model development, training pipelines, evaluation, and deployment. Use this rule to establish reproducible workflows, avoid common pitfalls like data leakage and poor evaluation metrics, and ensure models transition cleanly from development to production.
- Organize projects with separate data, model, training, and serving modules using tf.data pipelines
- Establish baseline models and validation strategies before scaling, with proper train/validation/test splits
- Track training with callbacks for checkpointing, early stopping, learning-rate scheduling, and TensorBoard logging
- Report task-appropriate metrics (AUROC, F1, confusion matrices, BLEU/ROUGE, MAE/RMSE) and evaluate on edge cases
- Export models with explicit input signatures and maintain preprocessing consistency between training and serving
- Identify common mistakes including data leakage, label quality issues, and batch/dtype mismatches between training and inference
Applies to
File patterns this rule matches.
Rule definition (reference)
Source of truth, from the repository.
TensorFlow and Deep Learning Rules
Project Structure
- Separate data loading, model definition, training, evaluation, and serving code.
- Use
tf.datapipelines for scalable input processing. - Keep model hyperparameters in typed config files or dataclasses.
- Store checkpoints, logs, and exported models outside source directories.
- Keep notebooks exploratory; move repeatable training code into modules.
Model Development
- Start with a small baseline model and a tiny overfit test before scaling.
- Use Keras layers and models unless lower-level TensorFlow APIs are required.
- Prefer explicit input shapes and named inputs/outputs.
- Use callbacks for checkpointing, early stopping, learning-rate scheduling, and TensorBoard logging.
- Use mixed precision only after validating numerical stability.
- Pin random seeds where reproducibility matters, while documenting nondeterministic GPU behavior.
Training
- Validate data shapes, dtypes, label ranges, and class balance before training.
- Split data before augmentation or normalization fitting.
- Use validation data for tuning and a separate test set for final reporting.
- Track loss curves, metrics, learning rate, and resource use.
- Save the best checkpoint by validation metric, not by final epoch.
Evaluation
- Report task-appropriate metrics such as AUROC, F1, calibration, perplexity, BLEU/ROUGE, or MAE/RMSE.
- Include confusion matrices or error slices for classification tasks.
- Evaluate on edge cases and distribution shifts when data allows.
- Compare against non-neural baselines when the dataset is small or tabular.
Deployment
- Export models with clear input signatures.
- Keep preprocessing consistent between training and serving.
- Add smoke tests that load the exported model and run inference on sample inputs.
- Monitor latency, memory, prediction drift, and input schema changes.
Common Mistakes
- Do not tune architecture before verifying labels and data quality.
- Do not leak validation data through preprocessing or augmentation.
- Do not rely on accuracy alone for imbalanced data.
- Do not deploy a notebook-only model.
- Do not ignore batch size, dtype, and device differences between training and inference.
Related rules
Senior full-stack TypeScript, React, Node.js guidance with clean architecture, testing, and WHY-oriented reasoning.
Quantitative factor research skills for designing, evaluating, and mining alpha factors in equities markets.
Android development with Jetpack Compose, clean architecture, and Material Design 3.
Angular development with Novo Elements UI library using standalone components.
Expert Angular 18 + TypeScript development with Jest, emphasizing clean code and performance.
Manage Kubernetes clusters, add-ons, stacks, and credentials via the Ankra CLI platform.