Community Picks·1y🤗 Transformers
🤗 Transformers provides APIs and tools for easily downloading and training state-of-the-art pretrained models for tasks in natural language processing, computer vision, audio, and multimodal categories. It supports interoperability between PyTorch, TensorFlow, and JAX, allowing for flexible model training and deployment. The library also offers comprehensive documentation, tutorials, and guides to help users get started and achieve specific goals.
Eli Bendersky·1yGoMLX: ML in Go without Python
The post describes implementing machine learning models in Go without using Python, focusing on the GoMLX package. GoMLX leverages XLA and PJRT for core functionalities and provides a Go-friendly interface. The author showcases re-implementations of ML inference tasks, such as a CNN model for the CIFAR-10 dataset and a more complex model like Gemma2, highlighting similar results to Python-based implementations.
DEV·1yHomemade GPT JS
A minimal re-implementation of Karpathy's minGPT using TensorFlow.js is introduced for educational purposes. This concise implementation, under 300 lines of code in TypeScript, allows users to train, experiment with parameters, and generate predictions directly in the browser using WebGPU or a Node.js environment. Various files including model.ts, config.ts, and trainer.ts are provided for detailed understanding and experimentation via a web or Node.js playground.