File size: 1,222 Bytes
0ac03fd aef0ec0 7aeddce aef0ec0 0ac03fd 7aeddce 79c6af6 4310f73 b9a3ea6 7aeddce 4310f73 79c6af6 b9a3ea6 79c6af6 b9a3ea6 7aeddce 79c6af6 7aeddce b9a3ea6 7aeddce b9a3ea6 7aeddce b9a3ea6 4310f73 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 |
---
license: apache-2.0
tags:
- unconditional-image-generation
datasets: skytnt/fbanimehq
---
# Model info
Project [fbanimegan](https://github.com/SkyTNT/fbanimegan)
### fbanime.pkl
StyleGan2 model trained with official [StyleGan3](https://github.com/NVlabs/stylegan3).
But I modified the code (networks_stylegan2.py and dataset.py) to support non-square resolutions.
FID: 1.4
### fbanime_fp32.pkl
fp32 version of fbanime.pkl
Note: Fp16 version (fbanime.pkl) only works on gpu. And fp32 version works on gpu and cpu.
### g_mapping.onnx
onnx format mapping network of fbanime_fp32.pkl
### g_synthesis.onnx
onnx format synthesis network of fbanime_fp32.pkl
### encoder.onnx
e4e model trained with [encoder4editing-stylegan3](https://github.com/yj7082126/encoder4editing-stylegan3).
I add support for official StyleGan2 model and change backbone to ResNet-34 in [restyle-encoder](https://github.com/yuval-alaluf/restyle-encoder).
### waifu_dect.onnx
YOLOv5 model trained with official [YOLOv5](https://github.com/ultralytics/yolov5)
# Usage
see [demo](https://huggingface.co/spaces/skytnt/full-body-anime-gan/blob/main/app.py)
# Dataset
[fbanimehq](https://huggingface.co/datasets/skytnt/fbanimehq) v2.0
|