BlinkDL/RWKV-LM
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.
PythonApache-2.0
Stargazers
- 1cvv
- aashiqmuhamedCarnegie Mellon University
- Albert-cordChina,ShenZhen
- C1-BA-B1-F3
- creke
- danyang-rainbowBeijing
- ddofer@SparkBeyond @LinialLab @Comocomo
- Evilran/dev/sd
- Hazukiaoi
- hongmiShanghai
- hwwh1999
- irplay
- jeffhsu3Ivy Natal
- jinmang2Republic of Korea
- lbr77@stevelbr
- ljmiao
- lvelveeIndie developer
- mahaoyangHangZhou China
- makeitt
- mook-wenyu
- phpcmsapp
- rkv0idBooking.com
- ruixu2Sun Yat-Sen University
- sgzwiz
- tomatobobotUnited States
- Vbansal21None
- Vivianluolita
- wangchichi1999
- WikidepiaIndonesia
- xieyan110
- yuulin
- zerclanzhang
- zhaoyun0071
- zielo-hueCalifornia
- zkvvv
- zylx0532