This model is a fine-tuned version of nvidia/segformer-b5-finetuned-ade-640-640 on the peldrak/grCoastline_512 dataset. It achieves the following results on the evaluation set:

Loss: 0.2332
Mean Iou: 0.6951
Mean Accuracy: 0.7738
Overall Accuracy: 0.9385
Accuracy Water: 0.9780
Accuracy Whitewater: 0.0
Accuracy Sediment: 0.8899
Accuracy Other Natural Terrain: 0.7486
Accuracy Vegetation: 0.9147
Accuracy Development: 0.8882
Accuracy Unknown: 0.9971
Iou Water: 0.9630
Iou Whitewater: 0.0
Iou Sediment: 0.7578
Iou Other Natural Terrain: 0.6446
Iou Vegetation: 0.8528
Iou Development: 0.6519
Iou Unknown: 0.9957
F1 Score: 0.9381

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: 4
eval_batch_size: 4
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Water	Accuracy Whitewater	Accuracy Sediment	Accuracy Other Natural Terrain	Accuracy Vegetation	Accuracy Development	Accuracy Unknown	Iou Water	Iou Whitewater	Iou Sediment	Iou Other Natural Terrain	Iou Vegetation	Iou Development	Iou Unknown	F1 Score
1.5035	0.24	20	1.3099	0.4529	0.5545	0.8166	0.7963	0.0007	0.7870	0.3478	0.8930	0.0604	0.9961	0.7606	0.0002	0.4736	0.2851	0.6104	0.0587	0.9819	0.8036
1.152	0.49	40	0.8933	0.4479	0.5546	0.8317	0.9700	0.0	0.9422	0.1145	0.8493	0.0126	0.9933	0.8446	0.0	0.5935	0.1113	0.5820	0.0126	0.9911	0.7930
1.3783	0.73	60	0.6157	0.5149	0.6084	0.8765	0.9728	0.0	0.9618	0.4069	0.9229	0.0020	0.9925	0.9109	0.0	0.5742	0.3812	0.7458	0.0020	0.9904	0.8560
0.8913	0.98	80	0.5418	0.5547	0.6332	0.8979	0.9783	0.0	0.9290	0.5524	0.9609	0.0173	0.9945	0.9338	0.0	0.6919	0.4900	0.7569	0.0173	0.9927	0.8800
0.9266	1.22	100	0.4282	0.5629	0.6421	0.9015	0.9816	0.0	0.9457	0.5864	0.9404	0.0447	0.9960	0.9308	0.0	0.6778	0.5128	0.7826	0.0443	0.9921	0.8855
0.6905	1.46	120	0.3734	0.6016	0.6758	0.9091	0.9885	0.0	0.8944	0.6737	0.9081	0.2712	0.9944	0.9338	0.0	0.6825	0.5550	0.7875	0.2598	0.9929	0.9022
0.7955	1.71	140	0.3400	0.6165	0.6850	0.9139	0.9705	0.0	0.8788	0.6373	0.9635	0.3481	0.9971	0.9335	0.0	0.7052	0.5782	0.7811	0.3233	0.9939	0.9077
0.5851	1.95	160	0.2810	0.6500	0.7179	0.9262	0.9749	0.0	0.9234	0.7170	0.9375	0.4775	0.9951	0.9487	0.0	0.7511	0.6320	0.8105	0.4145	0.9932	0.9230
0.4819	2.2	180	0.3615	0.5975	0.6868	0.8867	0.9779	0.0	0.9149	0.8533	0.6517	0.4151	0.9951	0.9454	0.0	0.7422	0.4986	0.6207	0.3815	0.9938	0.8866
0.8142	2.44	200	0.2637	0.6706	0.7485	0.9273	0.9840	0.0	0.8588	0.8254	0.8505	0.7243	0.9964	0.9556	0.0	0.7779	0.6342	0.7910	0.5411	0.9942	0.9276
0.396	2.68	220	0.2642	0.6508	0.7275	0.9186	0.9856	0.0	0.8634	0.7820	0.8452	0.6203	0.9958	0.9544	0.0	0.7546	0.5842	0.7684	0.4998	0.9943	0.9184
0.7178	2.93	240	0.2661	0.6598	0.7391	0.9222	0.9805	0.0	0.9402	0.7458	0.8377	0.6709	0.9988	0.9449	0.0	0.7474	0.5932	0.7888	0.5511	0.9934	0.9214
0.5699	3.17	260	0.2108	0.6815	0.7540	0.9326	0.9827	0.0	0.8490	0.7546	0.9139	0.7804	0.9975	0.9517	0.0	0.7857	0.6459	0.8094	0.5831	0.9947	0.9319
0.3768	3.41	280	0.2256	0.6891	0.7628	0.9363	0.9820	0.0	0.9426	0.7842	0.8824	0.7533	0.9952	0.9578	0.0	0.7728	0.6568	0.8343	0.6079	0.9941	0.9359
0.7524	3.66	300	0.2295	0.6758	0.7427	0.9314	0.9924	0.0	0.8485	0.7430	0.9262	0.6945	0.9944	0.9477	0.0	0.7486	0.6275	0.8293	0.5848	0.9929	0.9298
0.3174	3.9	320	0.2178	0.6780	0.7516	0.9319	0.9830	0.0	0.8744	0.7954	0.8833	0.7275	0.9979	0.9616	0.0	0.7345	0.6312	0.8332	0.5905	0.9953	0.9318
0.2191	4.15	340	0.2237	0.6714	0.7378	0.9311	0.9798	0.0	0.8692	0.8005	0.9113	0.6081	0.9954	0.9570	0.0	0.7618	0.6360	0.8232	0.5276	0.9944	0.9302
0.2526	4.39	360	0.2836	0.6616	0.7510	0.9192	0.9774	0.0	0.9351	0.7824	0.7882	0.7770	0.9973	0.9458	0.0	0.7342	0.5903	0.7671	0.5989	0.9951	0.9197
0.2428	4.63	380	0.2368	0.6873	0.7704	0.9316	0.9720	0.0	0.8632	0.8565	0.8365	0.8667	0.9981	0.9570	0.0	0.7836	0.6500	0.7979	0.6281	0.9948	0.9326
0.138	4.88	400	0.2237	0.6921	0.7655	0.9382	0.9765	0.0	0.8913	0.7857	0.9087	0.7983	0.9982	0.9618	0.0	0.7210	0.6680	0.8635	0.6355	0.9950	0.9382
0.2074	5.12	420	0.2075	0.6903	0.7617	0.9382	0.9855	0.0	0.8654	0.7844	0.9180	0.7821	0.9967	0.9614	0.0	0.7368	0.6624	0.8591	0.6175	0.9953	0.9377
0.4849	5.37	440	0.2016	0.6957	0.7713	0.9394	0.9873	0.0	0.8818	0.7709	0.9085	0.8538	0.9965	0.9497	0.0	0.7560	0.6707	0.8621	0.6362	0.9950	0.9390
0.2335	5.61	460	0.2293	0.6909	0.7638	0.9369	0.9848	0.0	0.8796	0.7226	0.9309	0.8326	0.9963	0.9568	0.0	0.7373	0.6366	0.8538	0.6566	0.9950	0.9358
0.2768	5.85	480	0.2453	0.6866	0.7644	0.9347	0.9764	0.0	0.9604	0.6923	0.9087	0.8168	0.9966	0.9562	0.0	0.7478	0.6367	0.8314	0.6390	0.9952	0.9335
0.1383	6.1	500	0.2322	0.6865	0.7570	0.9361	0.9547	0.0	0.8865	0.7609	0.9475	0.7510	0.9984	0.9443	0.0	0.7323	0.6723	0.8516	0.6097	0.9953	0.9354
0.3734	6.34	520	0.2335	0.6890	0.7657	0.9361	0.9877	0.0	0.8494	0.7954	0.8919	0.8377	0.9976	0.9586	0.0	0.7192	0.6598	0.8515	0.6386	0.9952	0.9360
0.131	6.59	540	0.2459	0.6849	0.7626	0.9362	0.9848	0.0	0.8292	0.7209	0.9426	0.8628	0.9981	0.9605	0.0	0.7203	0.6425	0.8589	0.6173	0.9951	0.9351
0.1874	6.83	560	0.2642	0.6761	0.7504	0.9308	0.9756	0.0	0.9512	0.7777	0.8681	0.6823	0.9975	0.9611	0.0	0.7349	0.6306	0.8235	0.5873	0.9954	0.9306
0.1282	7.07	580	0.2463	0.6883	0.7588	0.9380	0.9841	0.0	0.8800	0.7561	0.9309	0.7636	0.9970	0.9629	0.0	0.7370	0.6487	0.8647	0.6093	0.9954	0.9373
0.1173	7.32	600	0.2412	0.6898	0.7661	0.9374	0.9778	0.0	0.8615	0.7889	0.9097	0.8267	0.9981	0.9609	0.0	0.7282	0.6603	0.8610	0.6225	0.9954	0.9374
0.1012	7.56	620	0.2550	0.6869	0.7623	0.9353	0.9837	0.0	0.8689	0.7977	0.8907	0.7985	0.9968	0.9623	0.0	0.7376	0.6455	0.8458	0.6219	0.9951	0.9353
0.2309	7.8	640	0.2549	0.6920	0.7595	0.9386	0.9840	0.0	0.8390	0.7803	0.9350	0.7801	0.9979	0.9619	0.0	0.7304	0.6571	0.8609	0.6381	0.9956	0.9378
0.1986	8.05	660	0.2270	0.6979	0.7668	0.9405	0.9781	0.0	0.9063	0.7757	0.9203	0.7895	0.9980	0.9625	0.0	0.7509	0.6640	0.8620	0.6503	0.9955	0.9400
0.5972	8.29	680	0.2307	0.6997	0.7714	0.9404	0.9743	0.0	0.8955	0.7639	0.9248	0.8432	0.9983	0.9610	0.0	0.7591	0.6594	0.8578	0.6650	0.9953	0.9399
0.1218	8.54	700	0.2914	0.6734	0.7541	0.9223	0.9829	0.0	0.8578	0.8144	0.8106	0.8157	0.9974	0.9621	0.0	0.7434	0.5823	0.7703	0.6606	0.9950	0.9234
0.1051	8.78	720	0.2212	0.7029	0.7755	0.9422	0.9849	0.0	0.9237	0.7433	0.9202	0.8598	0.9969	0.9628	0.0	0.7848	0.6672	0.8522	0.6582	0.9953	0.9413
0.1204	9.02	740	0.2667	0.6915	0.7636	0.9368	0.9875	0.0	0.9098	0.6347	0.9566	0.8601	0.9967	0.9605	0.0	0.7803	0.6052	0.8308	0.6681	0.9955	0.9341
0.2844	9.27	760	0.2356	0.6913	0.7635	0.9352	0.9815	0.0	0.8728	0.7781	0.8930	0.8201	0.9990	0.9614	0.0	0.7466	0.6327	0.8340	0.6686	0.9956	0.9349
0.1134	9.51	780	0.2499	0.6941	0.7668	0.9372	0.9848	0.0	0.8818	0.7850	0.8968	0.8228	0.9968	0.9640	0.0	0.7495	0.6446	0.8432	0.6619	0.9958	0.9370
0.6034	9.76	800	0.2462	0.6932	0.7635	0.9371	0.9826	0.0	0.9008	0.7640	0.9119	0.7903	0.9949	0.9619	0.0	0.7692	0.6418	0.8382	0.6471	0.9944	0.9365
0.088	10.0	820	0.2285	0.7030	0.7754	0.9417	0.9862	0.0	0.9104	0.7763	0.9045	0.8533	0.9969	0.9643	0.0	0.7718	0.6665	0.8539	0.6686	0.9956	0.9412
0.0965	10.24	840	0.2332	0.6951	0.7738	0.9385	0.9780	0.0	0.8899	0.7486	0.9147	0.8882	0.9971	0.9630	0.0	0.7578	0.6446	0.8528	0.6519	0.9957	0.9381

Framework versions

Transformers 4.38.1
Pytorch 2.1.2
Datasets 2.18.0
Tokenizers 0.15.2

peldrak/segformer-b5-ade-finetuned-grCoastline

segformer-b5-ade-finetuned-grCoastline

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

peldrak/
segformer-b5-ade-finetuned-grCoastline