Instructions to use saberzl/So-Fake-R1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use saberzl/So-Fake-R1 with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
File size: 1,171 Bytes
405368d | 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 | {
"name": "So-Fake-R1",
"version": "v5.64",
"created": "2026-08-01",
"base_model": "Qwen/Qwen2.5-VL-7B-Instruct",
"components": {
"visual_observer": {
"type": "LoRA adapter",
"path": "observer_adapter",
"required_files": [
"adapter_config.json",
"adapter_model.safetensors",
"additional_config.json"
]
},
"forensic_evidence_provider": {
"type": "FETA provider weights, Qwen layer-19 feature source",
"path": "provider",
"required_files": [
"npr_internalized_feta_mono_blur_only05_e1.pt",
"feta_v0_L19_cls_only10d_L19_u1_e1.pt",
"npr_feta_cls_npr_residual_feta10d_e1.pt",
"l19_heatmap_bbox_head_best.pt"
]
},
"decision_maker": {
"type": "LoRA adapter",
"path": "decision_maker_adapter",
"required_files": [
"adapter_config.json",
"adapter_model.safetensors",
"additional_config.json"
]
}
},
"not_included": [
"Qwen2.5-VL-7B-Instruct base model",
"SAM2 checkpoints",
"optimizer states",
"scheduler states",
"merged observer checkpoint",
"training data"
]
}
|