finetune-instance-segmentation-ade20k-mini-mask2former

This model is a fine-tuned version of facebook/mask2former-swin-tiny-coco-instance on the yeray142/kitti-mots-instance dataset. It achieves the following results on the evaluation set:

• Loss: 21.4682
• Map: 0.2191
• Map 50: 0.4214
• Map 75: 0.2032
• Map Small: 0.1293
• Map Medium: 0.4299
• Map Large: 0.9458
• Mar 1: 0.0979
• Mar 10: 0.2731
• Mar 100: 0.3209
• Mar Small: 0.2542
• Mar Medium: 0.5212
• Mar Large: 0.9604
• Map Car: 0.406
• Mar 100 Car: 0.5312
• Map Person: 0.0323
• Mar 100 Person: 0.1106

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 10.0
• mixed_precision_training: Native AMP

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 Car	Mar 100 Car	Map Person	Mar 100 Person
32.2084	1.0	315	24.9517	0.1754	0.3241	0.1688	0.0901	0.3559	0.9036	0.0848	0.2249	0.2625	0.1937	0.4585	0.9383	0.3414	0.4833	0.0093	0.0417
26.4262	2.0	630	23.8438	0.1907	0.358	0.1793	0.1022	0.3764	0.9182	0.0904	0.2378	0.2851	0.2184	0.4779	0.9446	0.3657	0.5024	0.0157	0.0677
24.7264	3.0	945	22.7357	0.197	0.3715	0.189	0.1086	0.3803	0.9337	0.0912	0.2441	0.2901	0.2234	0.4819	0.9531	0.3769	0.5086	0.017	0.0716
23.7704	4.0	1260	22.5427	0.2001	0.3753	0.1878	0.1092	0.3902	0.9368	0.0924	0.2519	0.2994	0.2332	0.4914	0.9552	0.3791	0.513	0.0211	0.0858
22.7954	5.0	1575	22.0928	0.2071	0.3926	0.195	0.1184	0.4043	0.933	0.0961	0.2594	0.3075	0.2418	0.5028	0.9524	0.3906	0.5253	0.0237	0.0897
22.2719	6.0	1890	21.8539	0.2135	0.4034	0.1965	0.1216	0.4159	0.9446	0.0973	0.265	0.3128	0.2478	0.5031	0.9608	0.3985	0.5309	0.0285	0.0946
21.6338	7.0	2205	21.7856	0.2125	0.4048	0.1965	0.1201	0.4207	0.9388	0.0967	0.2641	0.3131	0.2466	0.5119	0.957	0.3956	0.524	0.0293	0.1023
21.3044	8.0	2520	21.4704	0.2152	0.4046	0.2003	0.1229	0.4233	0.9421	0.0983	0.2663	0.3149	0.2487	0.5109	0.9592	0.4002	0.5274	0.0301	0.1024
20.9003	9.0	2835	21.5561	0.2151	0.4079	0.1994	0.124	0.4264	0.946	0.0977	0.2678	0.3194	0.2535	0.5132	0.9598	0.3997	0.5286	0.0304	0.1102
20.5867	9.9698	3140	21.4682	0.2191	0.4214	0.2032	0.1293	0.4299	0.9458	0.0979	0.2731	0.3209	0.2542	0.5212	0.9604	0.406	0.5312	0.0323	0.1106

Framework versions

• Transformers 4.50.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0