segformer-b0-finetuned-segments-sidewalk-2

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

• Loss: 0.1287

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: 6e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 150

Training results

Training Loss	Epoch	Step	Validation Loss
2.9957	2.8571	20	3.4269
2.6593	5.7143	40	2.3621
1.9746	8.5714	60	1.2378
1.5998	11.4286	80	1.2329
1.3299	14.2857	100	0.8019
1.3781	17.1429	120	0.8478
2.1912	20.0	140	0.6386
1.0362	22.8571	160	0.6467
1.3817	25.7143	180	0.4496
0.8108	28.5714	200	0.4032
0.8187	31.4286	220	0.4650
0.6671	34.2857	240	0.3251
0.6062	37.1429	260	0.4035
1.4152	40.0	280	0.3076
1.3078	42.8571	300	0.2517
0.4267	45.7143	320	0.2405
0.5829	48.5714	340	0.2142
0.8742	51.4286	360	0.2055
0.3055	54.2857	380	0.2257
0.5966	57.1429	400	0.1559
0.5006	60.0	420	0.1927
0.4433	62.8571	440	0.1525
0.2377	65.7143	460	0.1597
0.2612	68.5714	480	0.1703
0.477	71.4286	500	0.1663
0.2006	74.2857	520	0.1427
0.2641	77.1429	540	0.1370
0.5154	80.0	560	0.1386
0.447	82.8571	580	0.1274
0.195	85.7143	600	0.1236
0.1643	88.5714	620	0.1420
0.4199	91.4286	640	0.1226
0.1644	94.2857	660	0.1419
0.312	97.1429	680	0.1365
0.3905	100.0	700	0.1181
0.4035	102.8571	720	0.1305
0.1411	105.7143	740	0.1262
0.3018	108.5714	760	0.1322
0.1332	111.4286	780	0.1317
0.303	114.2857	800	0.1205
0.2399	117.1429	820	0.1358
0.2488	120.0	840	0.1226
0.304	122.8571	860	0.1275
0.2278	125.7143	880	0.1280
0.2718	128.5714	900	0.1294
0.5304	131.4286	920	0.1320
0.1143	134.2857	940	0.1279
0.1075	137.1429	960	0.1258
0.2103	140.0	980	0.1349
0.1483	142.8571	1000	0.1230
0.287	145.7143	1020	0.1253
0.3606	148.5714	1040	0.1287

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.1+cpu
• Datasets 2.19.2
• Tokenizers 0.19.1