Verified record · from the answer index

rwkv-architecture

RNN+Transformer hybrid with O(n) inference. Linear time, infinite context, no KV cache. Train like GPT (parallel), infer like RNN (sequential). Linux Foundation AI project. Production at Windows, Office, NeMo. RWKV-7 (March 2025). Models up

skill:orchestra-research/ai-research-skills/rwkv-architecture · published by orchestra-research · source repository