enyac-group / QuambaLinks
The official repository of Quamba1 [ICLR 2025] & Quamba2 [ICML 2025]
β59Updated 3 months ago
Alternatives and similar repositories for Quamba
Users that are interested in Quamba are comparing it to the libraries listed below
Sorting:
- β142Updated 7 months ago
- HALO: Hadamard-Assisted Low-Precision Optimization and Training method for finetuning LLMs. π The official implementation of https://arxβ¦β20Updated 7 months ago
- LLM Inference with Microscaling Formatβ31Updated 10 months ago
- ShiftAddLLM: Accelerating Pretrained LLMs via Post-Training Multiplication-Less Reparameterizationβ111Updated 11 months ago
- β82Updated 8 months ago
- [ICLR 2025] Palu: Compressing KV-Cache with Low-Rank Projectionβ140Updated 7 months ago
- β33Updated last year
- xKV: Cross-Layer SVD for KV-Cache Compressionβ40Updated this week
- [EMNLP 2024] RoLoRA: Fine-tuning Rotated Outlier-free LLMs for Effective Weight-Activation Quantizationβ38Updated last year
- SQUEEZED ATTENTION: Accelerating Long Prompt LLM Inferenceβ53Updated 10 months ago
- β52Updated last year
- The evaluation framework for training-free sparse attention in LLMsβ96Updated 3 months ago
- [COLM 2025] Official PyTorch implementation of "Quantization Hurts Reasoning? An Empirical Study on Quantized Reasoning Models"β51Updated 2 months ago
- [ICML 2025] XAttention: Block Sparse Attention with Antidiagonal Scoringβ231Updated 2 months ago
- β29Updated 10 months ago
- β59Updated 2 months ago
- An efficient implementation of the NSA (Native Sparse Attention) kernelβ115Updated 3 months ago
- Fast and memory-efficient exact attentionβ69Updated 6 months ago
- GEAR: An Efficient KV Cache Compression Recipefor Near-Lossless Generative Inference of LLMβ168Updated last year
- 16-fold memory access reduction with nearly no lossβ105Updated 6 months ago
- β126Updated 3 months ago
- [ICML 2025] SparseLoRA: Accelerating LLM Fine-Tuning with Contextual Sparsityβ56Updated 2 months ago
- Code for paper: [ICLR2025 Oral] FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inferenceβ142Updated 4 months ago
- Fast Hadamard transform in CUDA, with a PyTorch interfaceβ233Updated 3 weeks ago
- Work in progress.β73Updated 2 months ago
- Transformers components but in Tritonβ34Updated 4 months ago
- An algorithm for weight-activation quantization (W4A4, W4A8) of LLMs, supporting both static and dynamic quantizationβ148Updated 4 months ago
- AFPQ code implementationβ23Updated last year
- This repository contains the training code of ParetoQ introduced in our work "ParetoQ Scaling Laws in Extremely Low-bit LLM Quantization"β103Updated 3 months ago
- Official Implementation of SLEB: Streamlining LLMs through Redundancy Verification and Elimination of Transformer Blocksβ37Updated 7 months ago