models

Learn to integrate TensorBoard with custom training loops in TensorFlow. Easily visualize scalars histograms and images to monitor your models effectively.

Learn how to save and load checkpoints with Orbit in TensorFlow models. Effortlessly manage training progress with automatic saves, restores, and pre-emption handling. Maximize your model training efficiency.

Learn efficient multi-device training with TensorFlow models. Discover how to configure CPU, GPU, and TPU setups using flag-driven strategies and factory functions for optimized performance.

Learn to implement masked language modeling pre-training using TensorFlow Models. Explore BertPretrainerV2, MaskedLM head layer, and MaskedLMTask for efficient model training.

Learn how to implement attention mechanisms in NLP models with TensorFlow Models. Discover reusable Keras-compatible attention layers for efficient model building.

Learn to preprocess images for vision models with TensorFlow Models. Discover a production-ready pipeline for normalization, geometric transforms, and photometric augmentation.

Learn how to use Non-Maximum Suppression NMS in TensorFlow Models vision pipelines. Optimize object detection with efficient batched NMS for thousands of proposals.

Implement video classification models using TensorFlow Models official/vision stack. Train on datasets like Kinetics-400 easily with 3D backbones and config-driven training.

Learn to use anchor generators for object detection in TensorFlow Models. Create bounding-box templates efficiently using protobuf configuration and the anchor generator builder for your models.

Implement image segmentation models using TensorFlow Models. Build networks, add decoders and heads, and configure losses with the official Vision library for efficient end-to-end training.

Implement object detection with RetinaNet using TensorFlow Models. Configure the RetinaNet dataclass, build the model, and train with RetinaNetTask for automatic focal loss, anchor generation, and NMS.

Learn how to configure hyperparameters for TensorFlow models in the tensorflow models repository. Easily override defaults for optimal performance.