lucidrains / ring-attention-pytorch
Implementation of π Ring Attention, from Liu et al. at Berkeley AI, in Pytorch
β451Updated last month
Related projects: β
- Transformers with Arbitrarily Large Contextβ613Updated last month
- Ring attention implementation with flash attentionβ529Updated this week
- Helpful tools and examples for working with flex-attentionβ341Updated last month
- A subset of PyTorch's neural network modules, written in Python using OpenAI's Triton.β452Updated last week
- [ICML 2024] CLLMs: Consistency Large Language Modelsβ337Updated last month
- Implementation of ST-Moe, the latest incarnation of MoE after years of research at Brain, in Pytorchβ278Updated 3 months ago
- Code for Adam-mini: Use Fewer Learning Rates To Gain More https://arxiv.org/abs/2406.16793β283Updated 2 weeks ago
- This repository contains the experimental PyTorch native float8 training UXβ210Updated last month
- Memory optimization and training recipes to extrapolate language models' context length to 1 million tokens, with minimal hardware.β606Updated last month
- Microsoft Automatic Mixed Precision Libraryβ505Updated 3 weeks ago
- Explorations into some recent techniques surrounding speculative decodingβ189Updated 11 months ago
- π Efficiently (pre)training foundation models with native PyTorch features, including FSDP for training and SDPA implementation of Flashβ¦β147Updated last week
- Code for exploring Based models from "Simple linear attention language models balance the recall-throughput tradeoff"β206Updated last month
- Official PyTorch implementation of QA-LoRAβ111Updated 6 months ago
- PyTorch implementation of Infini-Transformer from "Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attentionβ¦β271Updated 4 months ago
- Scalable toolkit for efficient model alignmentβ509Updated this week
- Code for paper: "QuIP: 2-Bit Quantization of Large Language Models With Guarantees"β339Updated 6 months ago
- For releasing code related to compression methods for transformers, accompanying our publicationsβ356Updated 2 weeks ago
- Official Implementation of EAGLE-1 and EAGLE-2β747Updated 3 weeks ago
- Triton-based implementation of Sparse Mixture of Experts.β166Updated 3 weeks ago
- [ICLR 2024] Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruningβ539Updated 6 months ago
- A family of compressed models obtained via pruning and knowledge distillationβ241Updated 3 weeks ago
- A repository for research on medium sized language models.β469Updated 3 weeks ago
- scalable and robust tree-based speculative decoding algorithmβ298Updated last month
- Efficient implementations of state-of-the-art linear attention models in Pytorch and Tritonβ1,190Updated this week
- A simple and effective LLM pruning approach.β614Updated last month
- [ICML 2024] Break the Sequential Dependency of LLM Inference Using Lookahead Decodingβ1,099Updated 7 months ago
- Repo for "Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture"β530Updated 3 months ago
- Annotated version of the Mamba paperβ445Updated 6 months ago
- Code for the paper "QMoE: Practical Sub-1-Bit Compression of Trillion-Parameter Models".β258Updated 10 months ago