segformer-b0-finetuned-segments-pv

This model is a fine-tuned version of nvidia/segformer-b0-finetuned-ade-512-512 on the mouadenna/satellite_PV_dataset_train_test dataset. It achieves the following results on the evaluation set:

• Loss: 0.0224
• Mean Iou: 0.8462
• Precision: 0.9229

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: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Precision
0.0043	1.0	3666	0.0095	0.7784	0.8863
0.0036	2.0	7332	0.0082	0.8127	0.8991
0.0004	3.0	10998	0.0085	0.7946	0.8844
0.0	4.0	14664	0.0082	0.8313	0.9130
0.0	5.0	18330	0.0089	0.8147	0.9092
0.002	6.0	21996	0.0117	0.8121	0.9275
0.0017	7.0	25662	0.0105	0.7984	0.8629
0.0	8.0	29328	0.0108	0.8169	0.8889
0.0029	9.0	32994	0.0133	0.8224	0.9096
0.006	10.0	36660	0.0106	0.8280	0.8829
0.026	11.0	40326	0.0102	0.8501	0.9210
0.0	12.0	43992	0.0118	0.8339	0.9022
0.0019	13.0	47658	0.0139	0.8360	0.9103
0.0018	14.0	51324	0.0140	0.8332	0.9161
0.0039	15.0	54990	0.0129	0.8297	0.9012
0.0025	16.0	58656	0.0166	0.8368	0.9030
0.0073	17.0	62322	0.0148	0.8334	0.8950
0.0017	18.0	65988	0.0157	0.8451	0.9166
0.0	19.0	69654	0.0184	0.8129	0.9161
0.0013	20.0	73320	0.0162	0.8333	0.9042
0.0014	21.0	76986	0.0167	0.8470	0.9178
0.0015	22.0	80652	0.0147	0.8429	0.9114
0.0015	23.0	84318	0.0149	0.8458	0.8978
0.0009	24.0	87984	0.0158	0.8416	0.9072
0.0014	25.0	91650	0.0144	0.8457	0.9185
0.0013	26.0	95316	0.0164	0.8482	0.9212
0.0043	27.0	98982	0.0162	0.8400	0.9005
0.0024	28.0	102648	0.0203	0.8468	0.9217
0.0	29.0	106314	0.0192	0.8431	0.9142
0.0	30.0	109980	0.0181	0.8477	0.9203
0.0	31.0	113646	0.0179	0.8484	0.9177
0.001	32.0	117312	0.0170	0.8485	0.9104
0.0007	33.0	120978	0.0184	0.8471	0.9113
0.0013	34.0	124644	0.0193	0.8487	0.9209
0.0016	35.0	128310	0.0169	0.8491	0.9182
0.0005	36.0	131976	0.0180	0.8476	0.9167
0.0016	37.0	135642	0.0212	0.8478	0.9239
0.0014	38.0	139308	0.0211	0.8455	0.9164
0.0	39.0	142974	0.0203	0.8468	0.9211
0.0	40.0	146640	0.0224	0.8462	0.9229

Framework versions

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