RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding.
Apache License 2.0
12.05k
stars
827
forks
source link
Adding README.md to `RWKV-v4` to redirect users to `RWKV-v4neo` #171
Even on the discord, it still seems like new users somehow lands here, and try to use this to train/finetune