Nota Fraud
notafraud
AI & ML interests
Llama.cpp-based inference and sampling trickery
Recent Activity
new activity
5 days ago
silveroxides/Chroma-LoRA-Experiments:Hyper-Chroma-low-step-LoRA is a strong stabilizer (v34)
new activity
22 days ago
Ewere/Qwen3-30B-A3B-abliterated-erotic:Thank you!
new activity
about 1 month ago
mlabonne/gemma-3-12b-it-abliterated:Refusals to jokes
Organizations
None yet
notafraud's activity
Hyper-Chroma-low-step-LoRA is a strong stabilizer (v34)
1
#3 opened 5 days ago
by
notafraud

Thank you!
1
#1 opened 22 days ago
by
notafraud

Refusals to jokes
#3 opened about 1 month ago
by
notafraud

In the right direction
🤝
1
3
#1 opened 2 months ago
by
notafraud

Imatrix file
5
#3 opened 3 months ago
by
notafraud

Context size clarification
3
#1 opened 3 months ago
by
notafraud

More censored than the original
4
#1 opened 4 months ago
by
notafraud

Hello, this is author of Illustrious series.
👍
🧠
18
15
#2 opened 4 months ago
by
AngelBottomless

Template tokens
2
#1 opened 4 months ago
by
notafraud

Why increase censorship?
21
#20 opened 4 months ago
by
notafraud

Formatting clarification
2
#1 opened 4 months ago
by
notafraud

Extra `<s>` in EOG and EOS.
3
#1 opened 6 months ago
by
notafraud

Great model!
❤️
2
2
#3 opened 7 months ago
by
notafraud

Template clarification.
6
#1 opened 7 months ago
by
notafraud

More censored than the base model
9
#1 opened 7 months ago
by
notafraud

More censored than the base model
9
#1 opened 7 months ago
by
notafraud

Template clarification.
6
#1 opened 7 months ago
by
notafraud

Template clarification.
6
#1 opened 7 months ago
by
notafraud

More censored than the base model
9
#1 opened 7 months ago
by
notafraud
