BlinkDL / AlbatrossLinks
Efficient RWKV inference engine. RWKV7 7.2B fp16 decoding 10250 tps @ single 5090.
☆71Updated last week
Alternatives and similar repositories for Albatross
Users that are interested in Albatross are comparing it to the libraries listed below
Sorting:
- RWKV-7: Surpassing GPT☆102Updated last year
- A large-scale RWKV v7(World, PRWKV, Hybrid-RWKV) inference. Capable of inference by combining multiple states(Pseudo MoE). Easy to deploy…☆46Updated 2 months ago
- RWKV in nanoGPT style☆197Updated last year
- https://x.com/BlinkDL_AI/status/1884768989743882276☆28Updated 7 months ago
- RWKV-LM-V7(https://github.com/BlinkDL/RWKV-LM) Under Lightning Framework☆52Updated this week
- A fast RWKV Tokenizer written in Rust☆54Updated 4 months ago
- The WorldRWKV project aims to implement training and inference across various modalities using the RWKV7 architecture. By leveraging diff…☆62Updated 2 weeks ago
- ☆39Updated 7 months ago
- RWKV, in easy to read code☆72Updated 9 months ago
- RWKV-X is a Linear Complexity Hybrid Language Model based on the RWKV architecture, integrating Sparse Attention to improve the model's l…☆53Updated 5 months ago
- ☆160Updated last week
- Course Project for COMP4471 on RWKV☆17Updated last year
- QuIP quantization☆61Updated last year
- An open source replication of the stawberry method that leverages Monte Carlo Search with PPO and or DPO☆29Updated 2 weeks ago
- Lightweight toolkit package to train and fine-tune 1.58bit Language models☆104Updated 7 months ago
- My Implementation of Q-Sparse: All Large Language Models can be Fully Sparsely-Activated☆33Updated last year
- A specialized RWKV-7 model for Othello(a.k.a. Reversi) that predicts legal moves, evaluates positions, and performs in-context search. It…☆43Updated 11 months ago
- Training-free Post-training Efficient Sub-quadratic Complexity Attention. Implemented with OpenAI Triton.☆148Updated last month
- A collection of tricks and tools to speed up transformer models☆194Updated last week
- RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best…☆59Updated 9 months ago
- MiSS is a novel PEFT method that features a low-rank structure but introduces a new update mechanism distinct from LoRA, achieving an exc…☆25Updated 2 months ago
- ☆66Updated 9 months ago
- Prepare for DeekSeek R1 inference: Benchmark CPU, DRAM, SSD, iGPU, GPU, ... with efficient code.☆73Updated 10 months ago
- ☆81Updated last year
- GoldFinch and other hybrid transformer components☆45Updated last year
- Tree Attention: Topology-aware Decoding for Long-Context Attention on GPU clusters☆131Updated last year
- An unofficial pytorch implementation of 'Efficient Infinite Context Transformers with Infini-attention'☆54Updated last year
- Fast modular code to create and train cutting edge LLMs☆68Updated last year
- Reinforcement Learning Toolkit for RWKV.(v6,v7,ARWKV) Distillation,SFT,RLHF(DPO,ORPO), infinite context training, Aligning. Exploring the…☆56Updated 3 months ago
- ☆34Updated 2 weeks ago