finetune-instance-segmentation-ade20k-mini-mask2former_augmentation

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: 21.8491
• Map: 0.2024
• Map 50: 0.3976
• Map 75: 0.1846
• Map Small: 0.1131
• Map Medium: 0.4171
• Map Large: 0.9371
• Mar 1: 0.098
• Mar 10: 0.2621
• Mar 100: 0.3113
• Mar Small: 0.2456
• Mar Medium: 0.5068
• Mar Large: 0.9545
• Map Car: 0.3761
• Mar 100 Car: 0.5206
• Map Person: 0.0288
• Mar 100 Person: 0.102

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
32.1236	1.0	315	25.0016	0.172	0.3235	0.1677	0.0881	0.3483	0.8944	0.0877	0.2218	0.2617	0.1935	0.4549	0.9347	0.3376	0.4775	0.0065	0.0459
25.8006	2.0	630	23.8844	0.1836	0.3505	0.1708	0.0973	0.3691	0.9132	0.09	0.2324	0.276	0.2074	0.4765	0.9434	0.3541	0.4901	0.0131	0.0619
24.098	3.0	945	23.1822	0.1892	0.3583	0.1751	0.0975	0.3823	0.9297	0.0904	0.2392	0.283	0.215	0.4814	0.9495	0.3616	0.4967	0.0168	0.0693
23.0237	4.0	1260	22.7127	0.1913	0.3692	0.1751	0.1017	0.3846	0.9289	0.0933	0.2437	0.289	0.2225	0.4827	0.9486	0.3635	0.5003	0.0191	0.0778
22.25	5.0	1575	22.5918	0.1933	0.3765	0.1754	0.1053	0.3951	0.9267	0.0934	0.2477	0.2916	0.2253	0.4829	0.9474	0.3648	0.5	0.0218	0.0832
21.7056	6.0	1890	21.9666	0.2019	0.3913	0.1833	0.1101	0.4037	0.9311	0.0965	0.256	0.2998	0.235	0.4911	0.9497	0.3775	0.5145	0.0263	0.0852
21.218	7.0	2205	22.1376	0.2002	0.3859	0.1841	0.1087	0.412	0.9299	0.0974	0.255	0.3003	0.2331	0.5004	0.9524	0.3751	0.5113	0.0254	0.0892
20.7151	8.0	2520	21.7431	0.2013	0.3953	0.1819	0.1105	0.411	0.9349	0.0973	0.2595	0.3059	0.2401	0.5016	0.9533	0.375	0.5178	0.0277	0.094
20.4197	9.0	2835	21.8546	0.2024	0.3925	0.184	0.1112	0.4136	0.9325	0.0971	0.2589	0.3044	0.2387	0.4965	0.9531	0.3781	0.5164	0.0267	0.0925
20.1339	9.9698	3140	21.8491	0.2024	0.3976	0.1846	0.1131	0.4171	0.9371	0.098	0.2621	0.3113	0.2456	0.5068	0.9545	0.3761	0.5206	0.0288	0.102

Framework versions

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