A unified, torch-native training framework for MTP and block-parallel speculative decoding.
☆284Jul 31, 2026Updated 3 months ago
Alternatives and similar repositories for AngelSpec
Users that are interested in AngelSpec are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- ☆17Sep 27, 2026Updated last month
- Block-R1: Multi-domain RL Post-training Benchmark for dLLMs.☆22Jul 30, 2026Updated 3 months ago
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆1,265Updated this week
- A PyTorch native library for training speculative decoding models☆267Oct 29, 2026Updated last week
- A unified library for building, evaluating, and storing speculative decoding algorithms for LLM inference in vLLM☆942Updated this week
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- The official repo of "Draft-OPD: On-Policy Distillation for Speculative Draft Models"☆84Sep 4, 2026Updated 2 months ago
- PyTorch code for our paper "Progressive Binarization with Semi-Structured Pruning for LLMs"☆12Jul 11, 2026Updated 3 months ago
- Model compression toolkit engineered for enhanced usability, comprehensiveness, and efficiency.☆1,784Oct 31, 2026Updated last week
- Official Inference Code of EAGLE-4☆1,157Updated this week
- ☆13Sep 29, 2026Updated last month
- DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms☆7,288Jul 9, 2026Updated 3 months ago
- MoonEP: A Perfectly Balanced Expert Parallelism Library via Dynamic Redundant Experts☆1,202Oct 12, 2026Updated 3 weeks ago
- LightningRL: Breaking the Accuracy–Parallelism Trade-off of Block-wise dLLMs via Reinforcement Learning☆33Apr 25, 2026Updated 6 months ago
- TokenSpeed is a speed-of-light LLM inference engine.☆2,877Updated this week
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- [ICCV 2025] QuantCache：Adaptive Importance-Guided Quantization with Hierarchical Latent and Layer Caching for Video Generation☆18Sep 26, 2025Updated last year
- ☆44Updated this week
- ☆85Updated this week
- [NeurIPS 2026] Free Draft-and-Verification: Toward Lossless Parallel Decoding for Diffusion Large Language Models☆24Sep 29, 2026Updated last month
- DFlash: Block Diffusion for Flash Speculative Decoding☆6,217Aug 18, 2026Updated 2 months ago
- Skills for writing tilelang and debugging with CUDA toolkits.☆149Updated this week
- DisCa: Accelerating Video Diffusion Transformers with Distillation-Compatible Learnable Feature Caching☆26Apr 15, 2026Updated 6 months ago
- ☆436Updated this week
- Collect LLM Quantization related papers, data, repositories☆12Aug 27, 2024Updated 2 years ago
- Managed Kubernetes at scale on DigitalOcean • AdDigitalOcean Kubernetes includes the control plane, bandwidth allowance, container registry, automatic updates, and more for free.
- Mixture-of-experts (MoE) training megakernel for NVL72s☆637Oct 1, 2026Updated last month
- [EMNLP2026 Main] Official implementation of “Domino: Decoupling Causal Modeling from Autoregressive Drafting in Speculative Decoding”.☆141Jul 25, 2026Updated 3 months ago
- ☆17Jul 13, 2026Updated 3 months ago
- FlashMemory DS-V4 Retriever: a lightweight retriever that sparsifies DeepSeek-V4 CSA KV-cache. Weights available on Hugging Face.☆111Jul 18, 2026Updated 3 months ago
- This repository collects Visual Autoregressive (VAR) modeling papers from 2024 to 2026 published at top-tier conferences, as well as rele…☆21Mar 13, 2026Updated 7 months ago
- ☆26Sep 26, 2026Updated last month
- Official implementation of "Learning To Draft: Adaptive Speculative Decoding with Reinforcement Learning" (ICLR 2026)☆24Mar 1, 2026Updated 8 months ago
- DeepSelect: TopK kernels for DeepSeek Sparse Attention (DSA) and Samplers☆533Sep 30, 2026Updated last month
- FlashKDA: high-performance Kimi Delta Attention kernels☆1,296Oct 26, 2026Updated last week
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- High Performance LLM Inference Operator Library☆1,189Sep 7, 2026Updated 2 months ago
- JetSpec++/JetSpec: Causal Parallel (Tree) Drafting with Self Refinement☆256Updated this week
- [ICML 2025] CommVQ: Commutative Vector Quantization for KV Cache Compression☆27Sep 2, 2025Updated last year
- Lightweight Benchmark Server for Agentic GPU Programming☆58Updated this week
- W4A16 MoE GEMM CUDA kernel. Tuned for H200, B200 & B300.☆14Oct 30, 2026Updated last week
- Efficient (Jet) 0-shot tuning-free context extension with strong performance (Long).☆13Sep 28, 2026Updated last month
- [ICML 2023] SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models☆11Dec 13, 2023Updated 2 years ago