segformer-b0-finetuned-segments-sidewalk-2

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

• Loss: 2.9327
• Mean Iou: 0.0763
• Mean Accuracy: 0.1260
• Overall Accuracy: 0.5923
• Per Category Iou: [nan, 0.15598158400203022, 0.6233750625153907, 0.0037560777123078824, 0.026995519273962765, 0.027599075064035524, 0.0, 0.0010671752114502803, 0.0, 0.0, 0.503652156236298, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.42226922942999406, 0.0, 0.0005751844669974061, 0.0, 0.0, 0.0, 0.015053303500921295, 0.0, 0.0, 0.0, 0.5380260834627074, 0.2004924888392474, 0.07113330974397604, 7.792680075848753e-05, 0.000328515111695138, 0.0025085129486024, 0.0]
• Per Category Accuracy: [nan, 0.17282441039529764, 0.9228726118961177, 0.00408103876916878, 0.028255152590055656, 0.029544523907019265, nan, 0.0010791707371488259, 0.0, 0.0, 0.8681646650418041, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.7122996003019028, 0.0, 0.0005801259615003622, 0.0, 0.0, nan, 0.02304960072549563, 0.0, 0.0, 0.0, 0.9348363685365858, 0.2596289024956107, 0.07122958643730157, 8.48216389425569e-05, 0.0005356047133214773, 0.0026059641588056346, 0.0]

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: 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: 0.05

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Per Category Iou	Per Category Accuracy
3.0624	0.03	10	3.1628	0.0726	0.1219	0.5758	[nan, 0.0878087898079964, 0.611982872765419, 0.0001999765816897758, 0.006930751650791711, 0.0208104329339671, 0.0, 0.0010631316774049914, 0.0, 0.0, 0.4839157481183621, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.39292052415275885, 0.0, 0.0003268797082673576, 0.0011424188270622699, 0.0, 0.0, 0.004317032040472175, 3.142508260307427e-05, 0.0, 0.0, 0.5537894233680722, 0.28184052017073197, 0.015966383939961543, 0.0002995587926924772, 0.0005713078253519804, 0.0035316933149879015, 0.0]	[nan, 0.09656561651317118, 0.9239613003877697, 0.00021265611687132485, 0.007163978434475801, 0.0222089828684614, nan, 0.0010774805715464, 0.0, 0.0, 0.8583517795809614, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.705533848895072, 0.0, 0.00033222625115695, 0.0011495555325644448, 0.0, nan, 0.008061062548807214, 3.244014792707455e-05, 0.0, 0.0, 0.8715627360179777, 0.3828074002074446, 0.01597238073499201, 0.0003298619292210546, 0.0011388100215281895, 0.003805890022240969, 0.0]
2.6259	0.05	20	2.9327	0.0763	0.1260	0.5923	[nan, 0.15598158400203022, 0.6233750625153907, 0.0037560777123078824, 0.026995519273962765, 0.027599075064035524, 0.0, 0.0010671752114502803, 0.0, 0.0, 0.503652156236298, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.42226922942999406, 0.0, 0.0005751844669974061, 0.0, 0.0, 0.0, 0.015053303500921295, 0.0, 0.0, 0.0, 0.5380260834627074, 0.2004924888392474, 0.07113330974397604, 7.792680075848753e-05, 0.000328515111695138, 0.0025085129486024, 0.0]	[nan, 0.17282441039529764, 0.9228726118961177, 0.00408103876916878, 0.028255152590055656, 0.029544523907019265, nan, 0.0010791707371488259, 0.0, 0.0, 0.8681646650418041, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.7122996003019028, 0.0, 0.0005801259615003622, 0.0, 0.0, nan, 0.02304960072549563, 0.0, 0.0, 0.0, 0.9348363685365858, 0.2596289024956107, 0.07122958643730157, 8.48216389425569e-05, 0.0005356047133214773, 0.0026059641588056346, 0.0]

Framework versions

• Transformers 4.18.0
• Pytorch 1.10.0+cu111
• Datasets 2.1.0
• Tokenizers 0.12.1