---
title: rwkv-4-world
canonical_url: "https://www.modelscope.cn/models/AI-ModelScope/rwkv-4-world"
md_url: "https://www.modelscope.cn/models/AI-ModelScope/rwkv-4-world.md"
repository: AI-ModelScope/rwkv-4-world
last_updated: 2023-07-10
license: "Apache License 2.0"
pipeline_tag: fill-mask
tasks:
  - fill-mask
library_name:
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 1340
stars: 16
tags:
  - pytorch
  - text-generation
  - causal-lm
  - rwkv
---

# rwkv-4-world

> rwkv-4-world - AI-ModelScope 在 ModelScope 开源的模型。RWKV-4 trained on 100+ world languages (70% English, 15% multilang, 15% code).

AI-ModelScope/rwkv-4-world 是 ModelScope 魔搭社区上的fill-mask模型，采用 Apache License 2.0 许可。

- **Repository**: AI-ModelScope/rwkv-4-world
- **License**: Apache License 2.0
- **Tasks**: fill-mask
- **Tags**: pytorch, text-generation, causal-lm, rwkv
- **Downloads**: 1340
- **Stars**: 16
- **Last updated**: 2023-07-10

Source: https://www.modelscope.cn/models/AI-ModelScope/rwkv-4-world

---

# RWKV-4 World

## Model Description

RWKV-4 trained on 100+ world languages (70% English, 15% multilang, 15% code).

Some_Pile + Some_RedPajama + Some_OSCAR + All_Wikipedia + All_ChatGPT_Data_I_can_find

How to use:
* use latest rwkv pip package (0.7.4+)
* use https://github.com/BlinkDL/ChatRWKV/blob/main/v2/benchmark_world.py to test it
* larger models are stronger even though not fully trained yet

The differences between World & Raven:
* set pipeline = PIPELINE(model, "rwkv_vocab_v20230424") instead of 20B_tokenizer.json (EXACTLY AS WRITTEN HERE. "rwkv_vocab_v20230424" is included in rwkv 0.7.4+)
* use Question/Answer or User/AI or Human/Bot for chat. **DO NOT USE Bob/Alice or Q/A**

For 0.1/0.4/1.5B models, use **fp32** for first layer (will overflow in fp16 at this moment - fixable in future), or bf16 if you have 30xx/40xx GPUs. Example strategy: cuda fp32 *1 -> cuda fp16

NOTE: the new greedy tokenizer (https://github.com/BlinkDL/ChatRWKV/blob/main/tokenizer/rwkv_tokenizer.py) will tokenize '\n\n' as one single token instead of ['\n','\n']

QA prompt (replace \n\n in xxx to \n):
```
Question: xxx

Answer:
```
and
```
Instruction: xxx

Input: xxx

Response:
```

A good chat prompt (replace \n\n in xxx to \n):
```
User: hi

Assistant: Hi. I am your assistant and I will provide expert full response in full details. Please feel free to ask any question and I will always answer it.

User: xxx

Assistant:
```