young-geng / mlxu
Machine Learning eXperiment Utilities
☆46Updated 9 months ago
Alternatives and similar repositories for mlxu:
Users that are interested in mlxu are comparing it to the libraries listed below
- Minimal but scalable implementation of large language models in JAX☆34Updated 4 months ago
- Transformer with Mu-Parameterization, implemented in Jax/Flax. Supports FSDP on TPU pods.☆30Updated 2 months ago
- A set of Python scripts that makes your experience on TPU better☆50Updated 8 months ago
- ☆37Updated 11 months ago
- A simple library for scaling up JAX programs☆134Updated 4 months ago
- some common Huggingface transformers in maximal update parametrization (µP)☆80Updated 2 years ago
- JAX implementation of the Mistral 7b v0.2 model☆35Updated 8 months ago
- A flexible and efficient implementation of Flash Attention 2.0 for JAX, supporting multiple backends (GPU/TPU/CPU) and platforms (Triton/…☆23Updated last week
- Language models scale reliably with over-training and on downstream tasks☆96Updated 11 months ago
- Simple and efficient pytorch-native transformer training and inference (batched)☆69Updated 11 months ago
- ☆20Updated last year
- Automatically take good care of your preemptible TPUs☆36Updated last year
- ☆51Updated 9 months ago
- If it quacks like a tensor...☆57Updated 3 months ago
- LoRA for arbitrary JAX models and functions☆135Updated last year
- Inference code for LLaMA models in JAX☆116Updated 9 months ago
- JAX bindings for Flash Attention v2☆87Updated 7 months ago
- A toolkit for scaling law research ⚖☆47Updated last month
- TPU pod commander is a package for managing and launching jobs on Google Cloud TPU pods.☆20Updated 8 months ago
- ☆47Updated last year
- ☆78Updated 10 months ago
- The simplest, fastest repository for training/finetuning medium-sized GPTs.☆99Updated 3 months ago
- ☆75Updated 8 months ago
- Train a SmolLM-style llm on fineweb-edu in JAX/Flax with an assortment of optimizers.☆17Updated this week
- Triton Implementation of HyperAttention Algorithm☆47Updated last year
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training☆122Updated 10 months ago
- ☆30Updated 3 months ago
- HomebrewNLP in JAX flavour for maintable TPU-Training☆48Updated last year
- Griffin MQA + Hawk Linear RNN Hybrid☆85Updated 10 months ago