Hugging Face's logo
BlinkDL's picture
38 1 31

BlinkDL

BlinkDL
pavelpps's profile picture Keron's profile picture BinaryWood's profile picture
·
https://www.rwkv.com/

AI & ML interests

RWKV is all you need

Recent Activity

updated a model 1 day ago
BlinkDL/temp-latest-training-models
updated a model 6 days ago
BlinkDL/rwkv-7-pile
posted an update 7 days ago
RWKV-7 "Goose" 0.4B trained w/ ctx4k automatically extrapolates to ctx32k+, and perfectly solves NIAH ctx16k 🤯 100% RNN and attention-free. Only trained on the Pile. No finetuning. Replicable training runs. tested by our community: https://github.com/Jellyfish042/LongMamba
View all activity

Organizations

RWKV's profile picture FreedomAI's profile picture rwkv-x-dev's profile picture Social Post Explorers's profile picture

Posts 3

view post
Post
1373
RWKV-7 "Goose" 0.4B trained w/ ctx4k automatically extrapolates to ctx32k+, and perfectly solves NIAH ctx16k 🤯 100% RNN and attention-free. Only trained on the Pile. No finetuning. Replicable training runs. tested by our community: https://github.com/Jellyfish042/LongMamba
view post
Post
3959
RWKV-6-world-v3 (+3.1T tokens) is our best multilingual 7B model as of now: BlinkDL/rwkv-6-world

It's 100% RNN and attention-free. MMLU 54.2% (previous world-v2.1 = 47.9%. note: without eval-boosting tricks such as annealing).

RWKV-7-world-v4 soon :)
View all posts