onnx-community/Qwen3-0.6B-ONNX
Text Generation • Updated • 18.8k • 46
AI & ML interests
None defined yet.
Recent Activity
View all activity
Xenova
updated 2
models 4 days ago
Xenova
published a
model 4 days ago
License, provenance and attribution for revision 2575352d61be1bf7225cf8f8b268a4678025fc58
1
#2 opened 16 days ago
by 
healingsoftly
Xenova
updated a
model 5 days ago
Felladrin
updated a
model 6 days ago
Felladrin
published a
model 6 days ago
osanseviero
authored a
paper 10 days ago
added response_template
#2 opened 19 days ago
by 
nico-martin
Redistribution terms and exact origin of latex_finetuned ONNX
3
#1 opened 21 days ago
by 
PPPPPkgg
Commercial redistribution license for Whisper Base ONNX
2
#8 opened 26 days ago
by 
JaditaSoftware
Xenova
updated a
model 13 days ago
Felladrin
updated a
model 15 days ago
Felladrin
published a
model 15 days ago
Felladrin
updated a
model 15 days ago
Felladrin
published a
model 15 days ago
Parveshiiii
posted an update 18 days ago
Post
2,796
Most AI image detectors work fine in the lab and fail on social media.
The reason is simple. Real photos get compressed, resized and re-uploaded again and again, and detectors trained on clean images read that damage as a sign of fakery. A real photo that has passed through a few platforms often gets flagged as AI-generated.
AIRealNet-2 is built to fix that. It is the upgrade to our original Modotte/AIRealNet, trained on images collected from platforms like Instagram and X, so it has seen the kinds of quality loss and compression real images go through online.
The training set has 98k images, split 50/50 between AI-generated and real, so the model has no class bias to lean on. Personal photos were excluded by design, including AI-edited ones.
The architecture is unchanged: a SwinV2 Tiny backbone with a binary classification head, about 0.2B parameters, working at 256x256 resolution. It stays fast and easy to deploy for research or production.
Model: Modotte/AIRealNet-2
The reason is simple. Real photos get compressed, resized and re-uploaded again and again, and detectors trained on clean images read that damage as a sign of fakery. A real photo that has passed through a few platforms often gets flagged as AI-generated.
AIRealNet-2 is built to fix that. It is the upgrade to our original Modotte/AIRealNet, trained on images collected from platforms like Instagram and X, so it has seen the kinds of quality loss and compression real images go through online.
The training set has 98k images, split 50/50 between AI-generated and real, so the model has no class bias to lean on. Personal photos were excluded by design, including AI-edited ones.
The architecture is unchanged: a SwinV2 Tiny backbone with a binary classification head, about 0.2B parameters, working at 256x256 resolution. It stays fast and easy to deploy for research or production.
Model: Modotte/AIRealNet-2
- 4 replies
Use close_pattern instead of strip_prefix/strip_suffix in response_template
#3 opened 18 days ago
by nico-martin