Downloads · 30 days
1.3K
34% of all-time downloads
BlinkDL/rwkv-5-world
rwkv-5-world is a text generation model from BlinkDL. Use it when you need the model to write or continue text. It is set up for rwkv. The card lists the license as apache-2.0.
Use rwkv pip package 0.8.22+ for RWKV-5 inference: https://pypi.org/project/rwkv/ (pipeline = PIPELINE(model, "rwkvvocabv20230424") for rwkv-world models)
Downloads · 30 days
1.3K
34% of all-time downloads
All-time downloads
3.7K
Public
Repo size
93.4 GB
Likes
272
Public
Click a slice to open those files.
.pth31.8 GB · 100%
From the Hugging Face model README
Use rwkv pip package 0.8.22+ for RWKV-5 inference: https://pypi.org/project/rwkv/ (pipeline = PIPELINE(model, "rwkv_vocab_v20230424") for rwkv-world models)
Online 7B Demo: https://huggingface.co/spaces/BlinkDL/RWKV-Gradio-2
Online 1.5B Demo: https://huggingface.co/spaces/BlinkDL/RWKV-Gradio-1
GUI: https://github.com/josStorer/RWKV-Runner (see Releases)
Convert to HF formet: https://github.com/BBuf/RWKV-World-HF-Tokenizer
For developer: https://github.com/BlinkDL/ChatRWKV/blob/main/API_DEMO_CHAT.py
https://github.com/BlinkDL/ChatRWKV/blob/main/RWKV_v5_demo.py
How it works: https://twitter.com/BlinkDL_AI/status/1685230712247795713
RWKV-5 trained on 100+ world languages (70% English, 15% multilang, 15% code).
World = Some_Pile + Some_SlimPajama + Some_StarCoder + Some_OSCAR + All_Wikipedia + All_ChatGPT_Data_I_can_find
RWKV-5 training: set --my_testing "r2r4" in latest RWKV-LM v4neo: https://github.com/BlinkDL/RWKV-LM
World v1 = 0.59T tokens
World v2 = 1.12T tokens
Imagine what happens when we use more data :)
Recommended fine-tuning format (use \n for newlines):
User: xxxxxxxxxxxxxxx
Assistant: xxxxxxxxxxxxxxx
xxxxxxxxxxxxxxx
xxxxxxxxxxxxxxx
User: xxxxxxxxxxxxxxx
xxxxxxxxxxxxxxx
Assistant: xxxxxxxxxxxxxxx
xxxxxxxxxxxxxxx
xxxxxxxxxxxxxxx
xxxxxxxxxxxxxxx
A good chat prompt (better replace \n\n in xxx to \n, such that there will be no newlines in xxx):
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:
QA prompt (better replace \n\n in xxx to \n, such that there will be no newlines in xxx):
Question: xxx
Answer:
and
Instruction: xxx
Input: xxx
Response:
!!! There should not be any space after your final ":" or you will upset the tokenizer and see non-English reponse !!!
!!! There should not be any space after your final ":" or you will upset the tokenizer and see non-English reponse !!!
!!! There should not be any space after your final ":" or you will upset the tokenizer and see non-English reponse !!!