haoliuhl / ringattention
Transformers with Arbitrarily Large Context
β637Updated 3 months ago
Related projects β
Alternatives and complementary repositories for ringattention
- Implementation of π Ring Attention, from Liu et al. at Berkeley AI, in Pytorchβ474Updated 2 weeks ago
- Ring attention implementation with flash attentionβ578Updated this week
- [ICML 2024] Break the Sequential Dependency of LLM Inference Using Lookahead Decodingβ1,141Updated 3 weeks ago
- Microsoft Automatic Mixed Precision Libraryβ522Updated last month
- [ICLR 2024] Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruningβ553Updated 8 months ago
- Official PyTorch implementation of QA-LoRAβ116Updated 7 months ago
- Memory optimization and training recipes to extrapolate language models' context length to 1 million tokens, with minimal hardware.β644Updated last month
- [NeurIPS 2024] KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantizationβ303Updated 2 months ago
- Helpful tools and examples for working with flex-attentionβ462Updated 2 weeks ago
- Scalable toolkit for efficient model alignmentβ614Updated this week
- [ICML 2024] CLLMs: Consistency Large Language Modelsβ351Updated this week
- Code for paper: "QuIP: 2-Bit Quantization of Large Language Models With Guarantees"β347Updated 8 months ago
- Explorations into some recent techniques surrounding speculative decodingβ210Updated last year
- USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inferenceβ353Updated last week
- π Efficiently (pre)training foundation models with native PyTorch features, including FSDP for training and SDPA implementation of Flashβ¦β191Updated 3 weeks ago
- Implementation of ST-Moe, the latest incarnation of MoE after years of research at Brain, in Pytorchβ291Updated 4 months ago
- Official repository for LightSeq: Sequence Level Parallelism for Distributed Training of Long Context Transformersβ196Updated 2 months ago
- Pipeline Parallelism for PyTorchβ725Updated 2 months ago
- [ICML 2024] SqueezeLLM: Dense-and-Sparse Quantizationβ643Updated 2 months ago
- FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.β611Updated 2 months ago
- β284Updated 7 months ago
- A simple and effective LLM pruning approach.β660Updated 3 months ago
- Triton-based implementation of Sparse Mixture of Experts.β184Updated last month
- Tutel MoE: An Optimized Mixture-of-Experts Implementationβ729Updated 2 weeks ago
- Official Implementation of EAGLE-1 (ICML'24) and EAGLE-2 (EMNLP'24)β817Updated last week
- distributed trainer for LLMsβ543Updated 5 months ago
- This repository contains the experimental PyTorch native float8 training UXβ211Updated 3 months ago
- PyTorch implementation of Infini-Transformer from "Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attentionβ¦β279Updated 6 months ago
- Minimalistic large language model 3D-parallelism trainingβ1,229Updated last week