segformer-b1-finetuned-segments-pv

This model is a fine-tuned version of nvidia/segformer-b1-finetuned-ade-512-512 on the mouadenna/satellite_PV_dataset_train_test dataset. It achieves the following results on the evaluation set:

• Loss: 0.0192
• Mean Iou: 0.8631
• Precision: 0.9304

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: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Precision
0.0051	1.0	3666	0.0084	0.8064	0.8864
0.0038	2.0	7332	0.0115	0.7607	0.9338
0.0001	3.0	10998	0.0089	0.8145	0.9115
0.0	4.0	14664	0.0078	0.8317	0.9063
0.0	5.0	18330	0.0093	0.8078	0.9244
0.0017	6.0	21996	0.0080	0.8370	0.9203
0.0019	7.0	25662	0.0085	0.8395	0.9163
0.0001	8.0	29328	0.0099	0.8379	0.8931
0.0019	9.0	32994	0.0100	0.8388	0.9225
0.0048	10.0	36660	0.0103	0.8422	0.9035
0.0238	11.0	40326	0.0132	0.8378	0.9169
0.0	12.0	43992	0.0093	0.8509	0.9254
0.0017	13.0	47658	0.0116	0.8417	0.9243
0.0014	14.0	51324	0.0127	0.8348	0.9017
0.0031	15.0	54990	0.0123	0.8463	0.9299
0.0016	16.0	58656	0.0109	0.8439	0.9062
0.0091	17.0	62322	0.0199	0.8143	0.9344
0.0017	18.0	65988	0.0155	0.8326	0.9184
0.0	19.0	69654	0.0128	0.8351	0.8971
0.0013	20.0	73320	0.0135	0.8360	0.8970
0.0015	21.0	76986	0.0151	0.8466	0.9055
0.0011	22.0	80652	0.0136	0.8525	0.9117
0.0016	23.0	84318	0.0129	0.8478	0.9052
0.0007	24.0	87984	0.0189	0.8422	0.9422
0.0012	25.0	91650	0.0134	0.8435	0.9070
0.0012	26.0	95316	0.0152	0.8532	0.9243
0.0028	27.0	98982	0.0145	0.8521	0.9273
0.0023	28.0	102648	0.0156	0.8566	0.9288
0.0	29.0	106314	0.0176	0.8494	0.9222
0.0	30.0	109980	0.0156	0.8542	0.9282
0.0	31.0	113646	0.0158	0.8578	0.9273
0.0012	32.0	117312	0.0171	0.8560	0.9258
0.0005	33.0	120978	0.0146	0.8534	0.9149
0.0016	34.0	124644	0.0199	0.8519	0.9250
0.0015	35.0	128310	0.0164	0.8559	0.9181
0.0005	36.0	131976	0.0164	0.8551	0.9176
0.0014	37.0	135642	0.0172	0.8594	0.9263
0.0008	38.0	139308	0.0178	0.8601	0.9273
0.0	39.0	142974	0.0153	0.8601	0.9281
0.0	40.0	146640	0.0165	0.8632	0.9324
0.0	41.0	150306	0.0172	0.8624	0.9328
0.0002	42.0	153972	0.0201	0.8590	0.9303
0.0033	43.0	157638	0.0180	0.8611	0.9347
0.0	44.0	161304	0.0155	0.8620	0.9283
0.0011	45.0	164970	0.0174	0.8624	0.9277
0.0004	46.0	168636	0.0192	0.8612	0.9316
0.0	47.0	172302	0.0185	0.8612	0.9232
0.0007	48.0	175968	0.0173	0.8623	0.9247
0.0007	49.0	179634	0.0196	0.8628	0.9295
0.0003	50.0	183300	0.0192	0.8631	0.9304

Framework versions

• Transformers 4.42.3
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1