Liuhong99 / Sophia
The official implementation of “Sophia: A Scalable Stochastic Second-order Optimizer for Language Model Pre-training”
☆958Updated last year
Alternatives and similar repositories for Sophia:
Users that are interested in Sophia are comparing it to the libraries listed below
- Effortless plugin and play Optimizer to cut model training costs by 50%. New optimizer that is 2x faster than Adam on LLMs.☆381Updated 10 months ago
- [NeurIPS 2023] MeZO: Fine-Tuning Language Models with Just Forward Passes. https://arxiv.org/abs/2305.17333☆1,099Updated last year
- Convolutions for Sequence Modeling☆877Updated 10 months ago
- maximal update parametrization (µP)☆1,493Updated 9 months ago
- Fast & Simple repository for pre-training and fine-tuning T5-style models☆1,000Updated 7 months ago
- Cramming the training of a (BERT-type) language model into limited compute.☆1,327Updated 10 months ago
- Implementation of MEGABYTE, Predicting Million-byte Sequences with Multiscale Transformers, in Pytorch☆640Updated 3 months ago
- minLoRA: a minimal PyTorch library that allows you to apply LoRA to any PyTorch model.☆457Updated last year
- Implementation of Recurrent Memory Transformer, Neurips 2022 paper, in Pytorch☆407Updated 3 months ago
- Code for the ALiBi method for transformer language models (ICLR 2022)☆521Updated last year
- Kernl lets you run PyTorch transformer models several times faster on GPU with a single line of code, and is designed to be easily hackab…☆1,564Updated last year
- Implementation of Memorizing Transformers (ICLR 2022), attention net augmented with indexing and retrieval of memories using approximate …☆633Updated last year
- Annotated version of the Mamba paper☆481Updated last year
- MLCommons Algorithmic Efficiency is a benchmark and competition measuring neural network training speedups due to algorithmic improvement…☆375Updated this week
- LOMO: LOw-Memory Optimization☆984Updated 9 months ago
- Language Modeling with the H3 State Space Model☆519Updated last year
- A Python-level JIT compiler designed to make unmodified PyTorch programs faster.☆1,038Updated 11 months ago
- Official code for ReLoRA from the paper Stack More Layers Differently: High-Rank Training Through Low-Rank Updates☆448Updated 11 months ago
- Implementation of the conditionally routed attention in the CoLT5 architecture, in Pytorch☆226Updated 7 months ago
- Helpful tools and examples for working with flex-attention☆720Updated this week
- ☆542Updated 4 months ago
- Repo for "Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture"☆549Updated 3 months ago
- An implementation of "Retentive Network: A Successor to Transformer for Large Language Models"☆1,182Updated last year
- Foundation Architecture for (M)LLMs☆3,068Updated last year
- Implementation of ST-Moe, the latest incarnation of MoE after years of research at Brain, in Pytorch☆327Updated 10 months ago
- A simple and effective LLM pruning approach.☆735Updated 8 months ago
- Implementation of RETRO, Deepmind's Retrieval based Attention net, in Pytorch☆861Updated last year
- For optimization algorithm research and development.☆505Updated this week
- Implementation of 💍 Ring Attention, from Liu et al. at Berkeley AI, in Pytorch☆510Updated 5 months ago
- Tutel MoE: Optimized Mixture-of-Experts Library, Support DeepSeek FP8/FP4☆800Updated this week