BlinkDL
BlinkDL
AI & ML interests
RWKV is all you need
Recent Activity
updated
a model
2 days ago
BlinkDL/temp-latest-training-models
updated
a model
7 days ago
BlinkDL/rwkv-7-pile
posted
an
update
8 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
Organizations
BlinkDL's activity
how is this model on common benchmarks
1
#2 opened 9 months ago
by
Yan0096
Max token discrepency
1
#8 opened 11 months ago
by
Junkemail7558
size of the vocabulary
1
#1 opened 11 months ago
by
yuimo
very disappointing censorship
1
#7 opened 11 months ago
by
awokeknowing
How to arrange texts for a role playing like chat?
3
#6 opened over 1 year ago
by
azige
Missing Config Files
2
#24 opened over 1 year ago
by
wagesj45
Strange Chinese answer for RWKV-4-Raven-7B-v12-Eng49%-Chn49%-Jpn1%-Other1%-20230530-ctx8192.pth
3
#23 opened over 1 year ago
by
Starlento
How to fine-tune this model?
4
#2 opened over 1 year ago
by
ZhangRC
Unable to load tokenizer
5
#1 opened over 1 year ago
by
ZhangRC
Loading the model into WebUI
1
#20 opened over 1 year ago
by
Respair
May I ask what is SC2016?
1
#2 opened over 1 year ago
by
Song1943
What is the correct parameters to finetune the 7B rwkv model
2
#19 opened over 1 year ago
by
fubincom
How to load local files using from_pretrained()?
1
#18 opened over 1 year ago
by
BeastyZ
Any plan for bigger model such as 30B?
9
#10 opened over 1 year ago
by
lpy86786
Self-reflection
2
#17 opened over 1 year ago
by
Raspbfox
RWKV used as model for intent classification followed by performing tasks on own (Similar to Auto-GPT).
4
#13 opened over 1 year ago
by
Parag09
Training an INT4 version of the 7B model
7
#14 opened over 1 year ago
by
Raspbfox
Amazing results with Raven 3B!! It speaks other languages, it knows the date.. How does this work?
5
#15 opened over 1 year ago
by
phi0112358
Japanese dataset
5
#1 opened over 1 year ago
by
Verah
You want to check out this fine tuning dataset
4
#16 opened over 1 year ago
by
KnutJaegersberg