File size: 1,391 Bytes
339f0ef 92d74a9 339f0ef d523cb8 339f0ef f8dfc57 339f0ef 56fd7b3 339f0ef 3524a7e 339f0ef |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 |
---
license: apache-2.0
---
This is a experimental model, yet it is the Most powerful RNN model in the world.
# Mobius RWKV r6 chat 12B 16k
Mobius is a RWKV v6 arch chat model, benifit from [Matrix-Valued States and Dynamic Recurrence](https://arxiv.org/abs/2404.05892)
## Introduction
Mobius is a RWKV v6 arch model, a state based RNN+CNN+Transformer Mixed language model pretrained on a certain amount of data.
In comparison with the previous released Mobius, the improvements include:
* Only 24G Vram to run this model locally with fp16;
* Significant performance improvement in chinese;
* Stable support of 16K context length.
* function call support ;
## Usage
Chat format: User: xxxx\n\nAssistant: xxx\n\n
Recommend Temp and topp: 1 0.3
function call format:
System: func xxxx
User: xxxx
Assistant: xxxx
Obersavtion: xxxx
Assistant: xxxx
## More details
Mobius 12B 16k based on RWKV v6 arch, which is leading state based RNN+CNN+Transformer Mixed large language model which focus opensouce community
* 10~100 trainning/inference cost reduce;
* state based,selected memory, which mean good at grok;
* community support.
## requirements
21.9G vram to run fp16, 13.7G for int8, 7.2 for nf4 with Ai00 server.
* [RWKV Runner](https://github.com/josStorer/RWKV-Runner)
* [Ai00 server](https://github.com/cgisky1980/ai00_rwkv_server)
## Benchmark
ceval 63.53
cmmlu 76.07
|