ruipeterpan / specreasonLinks
PoC for "SpecReason: Fast and Accurate Inference-Time Compute via Speculative Reasoning" [arXiv '25]
☆37Updated 2 weeks ago
Alternatives and similar repositories for specreason
Users that are interested in specreason are comparing it to the libraries listed below
Sorting:
- (ACL 2025 main) FR-Spec: Frequency-Ranked Speculative Sampling☆26Updated this week
- ☆28Updated 2 weeks ago
- [ICLR 2024 Spotlight] Code for the paper "Merge, Then Compress: Demystify Efficient SMoE with Hints from Its Routing Policy"☆84Updated 11 months ago
- [NeurIPS 2024] Fast Best-of-N Decoding via Speculative Rejection☆45Updated 7 months ago
- Official Implementation of SAM-Decoding: Speculative Decoding via Suffix Automaton☆28Updated 3 months ago
- ☆42Updated 6 months ago
- ☆50Updated 6 months ago
- Pytorch implementation of our paper accepted by ICML 2024 -- CaM: Cache Merging for Memory-efficient LLMs Inference☆39Updated 11 months ago
- ☆93Updated 2 weeks ago
- This repo contains the source code for: Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs☆38Updated 9 months ago
- TokenSkip: Controllable Chain-of-Thought Compression in LLMs☆145Updated 2 months ago
- [ICLR 2025] SWIFT: On-the-Fly Self-Speculative Decoding for LLM Inference Acceleration☆49Updated 3 months ago
- Code for paper: [ICLR2025 Oral] FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference☆107Updated 2 weeks ago
- The official implementation of paper: SimLayerKV: A Simple Framework for Layer-Level KV Cache Reduction.☆46Updated 7 months ago
- [ACL 2024] Not All Experts are Equal: Efficient Expert Pruning and Skipping for Mixture-of-Experts Large Language Models☆89Updated last year
- CoT-Valve: Length-Compressible Chain-of-Thought Tuning☆69Updated 3 months ago
- Source code for the paper "LongGenBench: Long-context Generation Benchmark"☆20Updated 7 months ago
- The Official Implementation of Ada-KV: Optimizing KV Cache Eviction by Adaptive Budget Allocation for Efficient LLM Inference☆76Updated 4 months ago
- [NeurIPS 2024] The official implementation of "Kangaroo: Lossless Self-Speculative Decoding for Accelerating LLMs via Double Early Exitin…☆57Updated 11 months ago
- LongSpec: Long-Context Speculative Decoding with Efficient Drafting and Verification☆53Updated 3 months ago
- 16-fold memory access reduction with nearly no loss☆94Updated 2 months ago
- The official implementation of the paper "Towards Efficient Mixture of Experts: A Holistic Study of Compression Techniques (TMLR)".☆70Updated 2 months ago
- ☆64Updated last month
- [ICML 2025] Reward-guided Speculative Decoding (RSD) for efficiency and effectiveness.☆31Updated last month
- ☆23Updated last week
- ☆49Updated last year
- [ICML'25] Our study systematically investigates massive values in LLMs' attention mechanisms. First, we observe massive values are concen…☆66Updated last week
- More Tokens, Lower Precision: Towards the Optimal Token-Precision Trade-off in KV Cache Compression☆11Updated 4 months ago
- ☆9Updated 8 months ago
- ☆18Updated 6 months ago