gitpulse

BlinkDL/RWKV-LM

RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RNN and transformer - great performance, linear time, constant space (no kv-cache), fast training, infinite ctx_len, and free sentence embedding.

14,778 sao1,027 forkPythonApache-2.0cập nhật 2026-11-04

Xem trên GitHub →

Đà tăng sao

Hôm nay+0 (+0.0%)
Tuần này+6 (+0.0%)
Tháng này+30 (+0.2%)

Chủ đề

attention-mechanismdeep-learninggptgpt-2gpt-3language-modellinear-attentionlstmpytorchrnntransformertransformersrwkvchatgpt