aitfindonesia/KomdigiUB-Gambling-Detection-RTDETR-R50

v3-rtdetr-r50-gambling-finetune

This model is a fine-tuned version of PekingU/rtdetr_r50vd_coco_o365 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 6.8580
• Map: 0.7154
• Map 50: 0.86
• Map 75: 0.7982
• Map Small: 0.4818
• Map Medium: 0.4823
• Map Large: 0.5059
• Mar 1: 0.6019
• Mar 10: 0.8463
• Mar 100: 0.876
• Mar Small: 0.8241
• Mar Medium: 0.8693
• Mar Large: 0.8723
• Map Banner Promo: 0.8704
• Mar 100 Banner Promo: 0.9604
• Map Cta Button: 0.7422
• Mar 100 Cta Button: 0.905
• Map Game Thumbnail: 0.6783
• Mar 100 Game Thumbnail: 0.9073
• Map Logo: 0.7334
• Mar 100 Logo: 0.848
• Map Menu Nav: 0.5527
• Mar 100 Menu Nav: 0.7593

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 300
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Small	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Small	Mar Medium	Mar Large	Map Banner Promo	Mar 100 Banner Promo	Map Cta Button	Mar 100 Cta Button	Map Game Thumbnail	Mar 100 Game Thumbnail	Map Logo	Mar 100 Logo	Map Menu Nav	Mar 100 Menu Nav
No log	1.0	107	20.5763	0.0922	0.1237	0.1082	0.1056	0.0185	0.2108	0.1468	0.331	0.4918	0.301	0.4038	0.5959	0.399	0.9118	0.0195	0.4808	0.0367	0.5013	0.0013	0.2806	0.0046	0.2846
No log	2.0	214	9.0251	0.5324	0.6626	0.6192	0.3219	0.4183	0.6377	0.462	0.7685	0.8228	0.7076	0.7962	0.8858	0.7688	0.9316	0.6739	0.8749	0.5019	0.8567	0.4994	0.7971	0.2179	0.6538
No log	3.0	321	7.7306	0.6155	0.7736	0.6929	0.3994	0.4662	0.7004	0.5238	0.796	0.8365	0.6789	0.8151	0.8954	0.831	0.9507	0.672	0.8703	0.6238	0.8584	0.6296	0.7878	0.3213	0.7154
No log	4.0	428	7.1577	0.6898	0.8521	0.7639	0.4178	0.5517	0.8467	0.5838	0.813	0.8497	0.7595	0.8232	0.9338	0.8822	0.9676	0.7291	0.8863	0.6569	0.8385	0.6631	0.8165	0.5179	0.7396
29.4032	5.0	535	6.8177	0.7202	0.8795	0.7981	0.4606	0.552	0.8828	0.5851	0.83	0.8689	0.7771	0.8316	0.9249	0.9107	0.9699	0.743	0.9005	0.6694	0.8797	0.665	0.8266	0.6131	0.7681
29.4032	6.0	642	6.4833	0.7517	0.9161	0.8315	0.4888	0.6263	0.9139	0.5966	0.8393	0.877	0.7423	0.8503	0.9425	0.9318	0.975	0.792	0.9114	0.721	0.8935	0.6903	0.8216	0.6235	0.7835
29.4032	7.0	749	6.7079	0.7452	0.9086	0.8268	0.4626	0.5953	0.9113	0.5999	0.8343	0.8751	0.7833	0.8383	0.9446	0.927	0.9691	0.7828	0.9087	0.7239	0.8987	0.6797	0.8144	0.6126	0.7846
29.4032	8.0	856	6.7679	0.743	0.8989	0.8209	0.4715	0.6187	0.9201	0.6016	0.8354	0.8729	0.7509	0.837	0.9499	0.9227	0.9699	0.78	0.91	0.7224	0.8909	0.6738	0.8158	0.616	0.778
29.4032	9.0	963	6.5281	0.7457	0.8999	0.8317	0.4659	0.6117	0.9221	0.5949	0.8343	0.8674	0.69	0.8338	0.9453	0.9264	0.9706	0.7669	0.9023	0.7098	0.8719	0.6927	0.8129	0.6326	0.7791
9.6433	10.0	1070	6.5610	0.7537	0.9065	0.8402	0.4672	0.6283	0.926	0.6029	0.8364	0.8735	0.7052	0.8386	0.9521	0.9314	0.9743	0.7786	0.9082	0.7256	0.8792	0.6937	0.8158	0.6392	0.7901

Framework versions

• Transformers 5.0.0.dev0
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1

BibTeX entry and citation info

@misc{lv2023detrs,
      title={DETRs Beat YOLOs on Real-time Object Detection},
      author={Yian Zhao and Wenyu Lv and Shangliang Xu and Jinman Wei and Guanzhong Wang and Qingqing Dang and Yi Liu and Jie Chen},
      year={2023},
      eprint={2304.08069},
      archivePrefix={arXiv},
      primaryClass={cs.CV}

@misc{rogge2025transformerstutorials,
  author       = {Rogge, Niels},
  title        = {Transformers Tutorials},
  year         = {2025},
  howpublished = {\url{https://github.com/NielsRogge/Transformers-Tutorials}}
}