rwkv-4-raven / README.md
BlinkDL's picture
Update README.md
370b3aa
|
raw
history blame
874 Bytes
metadata
language:
  - en
tags:
  - pytorch
  - text-generation
  - causal-lm
  - rwkv
license: apache-2.0
datasets:
  - the_pile

RWKV-4 "Raven"-series Models

Model Description

These are RWKV-4-Pile models 3B/7B/14B finetuned on Alpaca, CodeAlpaca, Guanaco, GPT4All, ShareGPT and more.

Gradio Demo: https://huggingface.co/spaces/BlinkDL/Raven-RWKV-7B

Use https://github.com/BlinkDL/ChatRWKV to run them.

See https://github.com/BlinkDL/RWKV-LM for details on the RWKV Language Model (100% RNN).

  • RWKV-4-Raven-Eng : 99% English + 1% Multilang
  • RWKV-4-Raven-EngAndMore : 97% English + 1.5% Chn Jpn + 1.5% Multilang
  • RWKV-4-Raven-EngChnJpn : 98% English + 1% Chn Jpn + 1% Multilang
  • RWKV-4-Raven-ChnEng : 49.5% English + 50% Chinese + 0.5% Multilang

Previous Raven models are in: https://huggingface.co/BlinkDL/rwkv-4-pile-7b https://huggingface.co/BlinkDL/rwkv-4-pile-14b