li-plus / flash-preferenceLinks
Accelerate LLM preference tuning via prefix sharing with a single line of code
☆42Updated last month
Alternatives and similar repositories for flash-preference
Users that are interested in flash-preference are comparing it to the libraries listed below
Sorting:
- Quantized Attention on GPU☆44Updated 8 months ago
- ☆50Updated 2 months ago
- Odysseus: Playground of LLM Sequence Parallelism☆72Updated last year
- ☆123Updated 2 months ago
- An efficient implementation of the NSA (Native Sparse Attention) kernel☆108Updated last month
- ☆75Updated 2 months ago
- Flash-Muon: An Efficient Implementation of Muon Optimizer☆149Updated last month
- A Suite for Parallel Inference of Diffusion Transformers (DiTs) on multi-GPU Clusters☆46Updated last year
- [ACL 2024] RelayAttention for Efficient Large Language Model Serving with Long System Prompts☆40Updated last year
- patches for huggingface transformers to save memory☆28Updated 2 months ago
- [ICML 2025] SparseLoRA: Accelerating LLM Fine-Tuning with Contextual Sparsity☆48Updated last month
- Code for paper: [ICLR2025 Oral] FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference☆124Updated 2 months ago
- Estimate MFU for DeepSeekV3☆25Updated 7 months ago
- ☆137Updated 5 months ago
- Transformers components but in Triton☆34Updated 2 months ago
- ☆44Updated last week
- GEAR: An Efficient KV Cache Compression Recipefor Near-Lossless Generative Inference of LLM☆165Updated last year
- ☆60Updated 3 months ago
- 16-fold memory access reduction with nearly no loss☆103Updated 4 months ago
- A curated list of recent papers on efficient video attention for video diffusion models, including sparsification, quantization, and cach…☆32Updated 2 weeks ago
- DeeperGEMM: crazy optimized version☆71Updated 3 months ago
- [CoLM'25] The official implementation of the paper <MoA: Mixture of Sparse Attention for Automatic Large Language Model Compression>☆141Updated 3 weeks ago
- Bridge Megatron-Core to Hugging Face/Reinforcement Learning☆67Updated this week
- [ICLR2025] Breaking Throughput-Latency Trade-off for Long Sequences with Speculative Decoding☆123Updated 8 months ago
- Vocabulary Parallelism☆19Updated 4 months ago
- An auxiliary project analysis of the characteristics of KV in DiT Attention.☆31Updated 8 months ago
- Framework to reduce autotune overhead to zero for well known deployments.☆79Updated last week
- Linear Attention Sequence Parallelism (LASP)☆85Updated last year
- HALO: Hadamard-Assisted Low-Precision Optimization and Training method for finetuning LLMs. 🚀 The official implementation of https://arx…☆18Updated 5 months ago
- PyTorch bindings for CUTLASS grouped GEMM.☆106Updated 2 months ago