segformer-b0-finetuned-pothole_0910_512

This model is a fine-tuned version of nvidia/mit-b0 on the alphaca/orthophoto_pothole_etc_0911 dataset. It achieves the following results on the evaluation set:

• Loss: 0.0060
• Mean Iou: 0.3613
• Mean Accuracy: 0.7225
• Overall Accuracy: 0.7225
• Accuracy Unlabeled: nan
• Accuracy Pothole: 0.7225
• Iou Unlabeled: 0.0
• Iou Pothole: 0.7225

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 300

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Pothole	Iou Unlabeled	Iou Pothole
0.1419	5.5556	100	0.1363	0.0	0.0	0.0	nan	0.0	0.0	0.0
0.0483	11.1111	200	0.0476	0.0	0.0	0.0	nan	0.0	0.0	0.0
0.0264	16.6667	300	0.0396	0.0002	0.0005	0.0005	nan	0.0005	0.0	0.0005
0.025	22.2222	400	0.0196	0.0201	0.0401	0.0401	nan	0.0401	0.0	0.0401
0.0136	27.7778	500	0.0202	0.0886	0.1771	0.1771	nan	0.1771	0.0	0.1771
0.0118	33.3333	600	0.0201	0.1940	0.3880	0.3880	nan	0.3880	0.0	0.3880
0.0087	38.8889	700	0.0132	0.2738	0.5476	0.5476	nan	0.5476	0.0	0.5476
0.0115	44.4444	800	0.0116	0.2922	0.5843	0.5843	nan	0.5843	0.0	0.5843
0.0061	50.0	900	0.0099	0.2666	0.5331	0.5331	nan	0.5331	0.0	0.5331
0.0049	55.5556	1000	0.0094	0.3046	0.6091	0.6091	nan	0.6091	0.0	0.6091
0.0038	61.1111	1100	0.0087	0.3494	0.6989	0.6989	nan	0.6989	0.0	0.6989
0.0056	66.6667	1200	0.0073	0.3532	0.7063	0.7063	nan	0.7063	0.0	0.7063
0.0045	72.2222	1300	0.0083	0.3509	0.7019	0.7019	nan	0.7019	0.0	0.7019
0.0064	77.7778	1400	0.0070	0.3438	0.6875	0.6875	nan	0.6875	0.0	0.6875
0.0067	83.3333	1500	0.0068	0.3266	0.6533	0.6533	nan	0.6533	0.0	0.6533
0.0045	88.8889	1600	0.0066	0.3497	0.6995	0.6995	nan	0.6995	0.0	0.6995
0.004	94.4444	1700	0.0065	0.3421	0.6841	0.6841	nan	0.6841	0.0	0.6841
0.0031	100.0	1800	0.0071	0.3749	0.7498	0.7498	nan	0.7498	0.0	0.7498
0.0052	105.5556	1900	0.0062	0.3708	0.7416	0.7416	nan	0.7416	0.0	0.7416
0.0025	111.1111	2000	0.0060	0.3602	0.7205	0.7205	nan	0.7205	0.0	0.7205
0.003	116.6667	2100	0.0060	0.3669	0.7338	0.7338	nan	0.7338	0.0	0.7338
0.0032	122.2222	2200	0.0059	0.3756	0.7512	0.7512	nan	0.7512	0.0	0.7512
0.0023	127.7778	2300	0.0058	0.3579	0.7158	0.7158	nan	0.7158	0.0	0.7158
0.0035	133.3333	2400	0.0058	0.3609	0.7218	0.7218	nan	0.7218	0.0	0.7218
0.0017	138.8889	2500	0.0056	0.3464	0.6928	0.6928	nan	0.6928	0.0	0.6928
0.0042	144.4444	2600	0.0060	0.3532	0.7064	0.7064	nan	0.7064	0.0	0.7064
0.0052	150.0	2700	0.0057	0.3606	0.7212	0.7212	nan	0.7212	0.0	0.7212
0.0033	155.5556	2800	0.0057	0.3516	0.7032	0.7032	nan	0.7032	0.0	0.7032
0.0018	161.1111	2900	0.0057	0.3614	0.7228	0.7228	nan	0.7228	0.0	0.7228
0.003	166.6667	3000	0.0060	0.3784	0.7567	0.7567	nan	0.7567	0.0	0.7567
0.003	172.2222	3100	0.0057	0.3635	0.7270	0.7270	nan	0.7270	0.0	0.7270
0.0027	177.7778	3200	0.0055	0.3653	0.7307	0.7307	nan	0.7307	0.0	0.7307
0.0014	183.3333	3300	0.0057	0.3537	0.7075	0.7075	nan	0.7075	0.0	0.7075
0.0017	188.8889	3400	0.0056	0.3695	0.7389	0.7389	nan	0.7389	0.0	0.7389
0.0015	194.4444	3500	0.0057	0.3736	0.7473	0.7473	nan	0.7473	0.0	0.7473
0.0029	200.0	3600	0.0057	0.3561	0.7122	0.7122	nan	0.7122	0.0	0.7122
0.0035	205.5556	3700	0.0057	0.3558	0.7115	0.7115	nan	0.7115	0.0	0.7115
0.003	211.1111	3800	0.0058	0.3479	0.6958	0.6958	nan	0.6958	0.0	0.6958
0.0024	216.6667	3900	0.0057	0.3694	0.7388	0.7388	nan	0.7388	0.0	0.7388
0.0027	222.2222	4000	0.0057	0.3620	0.7240	0.7240	nan	0.7240	0.0	0.7240
0.0019	227.7778	4100	0.0058	0.3653	0.7306	0.7306	nan	0.7306	0.0	0.7306
0.0015	233.3333	4200	0.0057	0.3691	0.7381	0.7381	nan	0.7381	0.0	0.7381
0.0008	238.8889	4300	0.0059	0.3573	0.7146	0.7146	nan	0.7146	0.0	0.7146
0.002	244.4444	4400	0.0061	0.3475	0.6950	0.6950	nan	0.6950	0.0	0.6950
0.0013	250.0	4500	0.0060	0.3583	0.7167	0.7167	nan	0.7167	0.0	0.7167
0.0021	255.5556	4600	0.0060	0.3515	0.7031	0.7031	nan	0.7031	0.0	0.7031
0.0025	261.1111	4700	0.0061	0.3550	0.7100	0.7100	nan	0.7100	0.0	0.7100
0.0032	266.6667	4800	0.0060	0.3588	0.7176	0.7176	nan	0.7176	0.0	0.7176
0.0021	272.2222	4900	0.0059	0.3635	0.7270	0.7270	nan	0.7270	0.0	0.7270
0.0008	277.7778	5000	0.0060	0.3659	0.7319	0.7319	nan	0.7319	0.0	0.7319
0.0013	283.3333	5100	0.0059	0.3662	0.7325	0.7325	nan	0.7325	0.0	0.7325
0.0011	288.8889	5200	0.0060	0.3589	0.7179	0.7179	nan	0.7179	0.0	0.7179
0.0021	294.4444	5300	0.0060	0.3590	0.7179	0.7179	nan	0.7179	0.0	0.7179
0.0023	300.0	5400	0.0060	0.3613	0.7225	0.7225	nan	0.7225	0.0	0.7225

Framework versions

• Transformers 4.42.3
• Pytorch 2.3.1+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1