sail-sg / LongSpecLinks
LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification
☆65Updated 3 months ago
Alternatives and similar repositories for LongSpec
Users that are interested in LongSpec are comparing it to the libraries listed below
Sorting:
- ☆119Updated 4 months ago
- The official implementation of paper: SimLayerKV: A Simple Framework for Layer-Level KV Cache Reduction.☆49Updated last year
- [NeurIPS 2024] Fast Best-of-N Decoding via Speculative Rejection☆52Updated 11 months ago
- The official implementation for [NeurIPS2025 Oral] Gated Attention for Large Language Models: Non-linearity, Sparsity, and Attention-Sink…☆95Updated last month
- Official Implementation of SAM-Decoding: Speculative Decoding via Suffix Automaton☆33Updated 8 months ago
- Implementation for FP8/INT8 Rollout for RL training without performence drop.☆261Updated 3 weeks ago
- Source code for the paper "LongGenBench: Long-context Generation Benchmark"☆23Updated last year
- ☆44Updated 5 months ago
- PoC for "SpecReason: Fast and Accurate Inference-Time Compute via Speculative Reasoning" [NeurIPS '25]☆56Updated 3 weeks ago
- This repo contains the source code for: Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs☆40Updated last year
- [ICLR‘24 Spotlight] Code for the paper "Merge, Then Compress: Demystify Efficient SMoE with Hints from Its Routing Policy"☆96Updated 4 months ago
- [ACL 2025] Squeezed Attention: Accelerating Long Prompt LLM Inference☆54Updated 11 months ago
- Pytorch implementation of our paper accepted by ICML 2024 -- CaM: Cache Merging for Memory-efficient LLMs Inference☆47Updated last year
- ☆61Updated 3 months ago
- ☆104Updated last month
- ☆10Updated last year
- [ICLR 2025] SWIFT: On-the-Fly Self-Speculative Decoding for LLM Inference Acceleration☆55Updated 8 months ago
- ☆53Updated last year
- The Official Implementation of Ada-KV [NeurIPS 2025]☆106Updated last month
- Kinetics: Rethinking Test-Time Scaling Laws☆81Updated 3 months ago
- The official implementation of the paper "Towards Efficient Mixture of Experts: A Holistic Study of Compression Techniques (TMLR)".☆79Updated 7 months ago
- ☆38Updated last year
- Multi-Candidate Speculative Decoding☆36Updated last year
- 16-fold memory access reduction with nearly no loss☆105Updated 7 months ago
- Ouroboros: Speculative Decoding with Large Model Enhanced Drafting (EMNLP 2024 main)☆110Updated 7 months ago
- ☆98Updated last month
- Code for the preprint "Cache Me If You Can: How Many KVs Do You Need for Effective Long-Context LMs?"☆46Updated 2 months ago
- [ICLR 2025] PEARL: Parallel Speculative Decoding with Adaptive Draft Length☆120Updated 6 months ago
- ☆46Updated 11 months ago
- ☆29Updated 5 months ago