li-plus / flash-preferenceLinks
Accelerate LLM preference tuning via prefix sharing with a single line of code
☆41Updated 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 6 months ago
- ☆49Updated 2 weeks ago
- Odysseus: Playground of LLM Sequence Parallelism☆69Updated 11 months ago
- ☆93Updated last week
- ☆70Updated 2 weeks ago
- ☆59Updated last month
- 16-fold memory access reduction with nearly no loss☆94Updated 2 months ago
- Transformers components but in Triton☆33Updated 3 weeks ago
- DeeperGEMM: crazy optimized version☆69Updated last month
- An auxiliary project analysis of the characteristics of KV in DiT Attention.☆30Updated 6 months ago
- [ICLR2025] Breaking Throughput-Latency Trade-off for Long Sequences with Speculative Decoding☆116Updated 6 months ago
- Beyond KV Caching: Shared Attention for Efficient LLMs☆19Updated 10 months ago
- A Suite for Parallel Inference of Diffusion Transformers (DiTs) on multi-GPU Clusters☆45Updated 10 months ago
- Flash-Muon: An Efficient Implementation of Muon Optimizer☆121Updated last week
- ☆21Updated 2 months ago
- The official implementation of paper: SimLayerKV: A Simple Framework for Layer-Level KV Cache Reduction.☆46Updated 7 months ago
- Official implementation of "The Sparse Frontier: Sparse Attention Trade-offs in Transformer LLMs"☆32Updated last month
- Estimate MFU for DeepSeekV3☆24Updated 5 months ago
- ☆76Updated last month
- ☆129Updated 3 months ago
- [ACL 2024] RelayAttention for Efficient Large Language Model Serving with Long System Prompts☆39Updated last year
- Code for paper: [ICLR2025 Oral] FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference☆107Updated 2 weeks ago
- FastCache: Fast Caching for Diffusion Transformer Through Learnable Linear Approximation [Efficient ML Model]☆21Updated this week
- Best practices for testing advanced Mixtral, DeepSeek, and Qwen series MoE models using Megatron Core MoE.☆17Updated this week
- SQUEEZED ATTENTION: Accelerating Long Prompt LLM Inference☆46Updated 6 months ago
- [ICLR 2024] Jaiswal, A., Gan, Z., Du, X., Zhang, B., Wang, Z., & Yang, Y. Compressing llms: The truth is rarely pure and never simple.☆24Updated last month
- Vocabulary Parallelism☆19Updated 2 months ago
- [ICLR 2025] DeFT: Decoding with Flash Tree-attention for Efficient Tree-structured LLM Inference☆23Updated last month
- XAttention: Block Sparse Attention with Antidiagonal Scoring☆158Updated 3 weeks ago
- Linear Attention Sequence Parallelism (LASP)☆83Updated last year