xllm: GitHub releases light-weight infrastructure for express LLM pre-training

Litelm: LiteLLM Without the Bloat

Releases·Tool Release

OpenArch – PyTorch implementations of modern LLM architectures

OpenArch: PyTorch implementations of modern LLM architectures released on GitHub

OpenArch: PyTorch implementations of modern LLM architectures released on GitHub

Horus-runtime is a tool that lets you train a small language model from scratch, aimed at developers and researchers who want to understand LLM training or build custom small models. It provides a runtime environment for training, likely with a focus on simplicity and educational value.

A new interactive web essay titled 'The Next Token — LLMs, from the beginning' has been published on ceres1.space, offering a comprehensive introduction to large language models. The resource appears aimed at developers and enthusiasts seeking a foundational understanding of LLMs.

Products·launched

Text LLM training from scratch with PyTorch

A clean, readable codebase that implements the full LLM training pipeline (pretraining, SFT, DPO, GRPO/RL) using only PyTorch primitives, avoiding high-level abstractions. It helps developers understand the underlying math and mechanics of LLM training.

A clean, readable codebase that implements the full LLM training pipeline (pretraining, SFT, DPO, GRPO/RL) using only PyTorch primitives, avoiding high-level abstractions. It helps developers understand the underlying math and mechanics of LLM training.