segformer-b5-ade-finetuned-coastTrain

peldrak/segformer-b5-ade-finetuned-coastTrain

segformer-b5-ade-finetuned-coastTrain is a image segmentation model from peldrak. Use it for the image segmentation task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as other.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

10% of all-time downloads

All-time downloads

Public

Parameters

84.6M

339 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors339 MB · 100%

At a glance

Task: Image Segmentation
Library: transformers
License: other
Model type: segformer
Access: Public
Created: Mar 6, 2024
Updated: Mar 6, 2024
SHA: 07ab89dd

Base models

nvidia/segformer-b5-finetuned-ade-640-640

Model card

From the Hugging Face model README

segformer-b5-ade-finetuned-coastTrain

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

Loss: 0.3412
Mean Iou: 0.7737
Mean Accuracy: 0.8545
Overall Accuracy: 0.9252
Accuracy Water: 0.9680
Accuracy Whitewater: 0.6271
Accuracy Sediment: 0.8745
Accuracy Other Natural Terrain: 0.8030
Accuracy Vegetation: 0.9108
Accuracy Development: 0.8296
Accuracy Unknown: 0.9682
Iou Water: 0.9285
Iou Whitewater: 0.5294
Iou Sediment: 0.8047
Iou Other Natural Terrain: 0.6748
Iou Vegetation: 0.8430
Iou Development: 0.7686
Iou Unknown: 0.8673
F1 Score: 0.9245

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: 20

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
1.6133	0.16	20	1.4728	0.4356	0.5582	0.7600	0.8440	0.0000	0.5434	0.2624	0.8192	0.5884	0.8502	0.7387	0.0000	0.4669	0.1171	0.5786	0.4642	0.6833	0.7532
1.4431	0.31	40	1.0911	0.4768	0.5735	0.8206	0.9076	0.0	0.7765	0.0006	0.9006	0.6686	0.7608	0.8394	0.0	0.5832	0.0006	0.6596	0.4954	0.7597	0.8114
1.5275	0.47	60	0.7874	0.5222	0.5966	0.8585	0.9553	0.0	0.7752	0.0	0.9101	0.6629	0.8727	0.8565	0.0	0.6318	0.0	0.7222	0.5763	0.8688	0.8454
1.1583	0.62	80	0.7113	0.5128	0.6001	0.8440	0.8830	0.0	0.8460	0.0	0.9323	0.6588	0.8809	0.8283	0.0	0.5762	0.0	0.7216	0.5873	0.8762	0.8348
0.9739	0.78	100	0.6059	0.5450	0.6229	0.8733	0.9534	0.0	0.8246	0.0	0.8876	0.8146	0.8803	0.8703	0.0	0.6770	0.0	0.7506	0.6408	0.8760	0.8615
0.838	0.93	120	0.5250	0.5557	0.6273	0.8800	0.9679	0.0	0.8549	0.0	0.8665	0.8087	0.8934	0.8677	0.0	0.6840	0.0	0.7703	0.6847	0.8830	0.8674
0.5815	1.09	140	0.4973	0.5528	0.6247	0.8788	0.9630	0.0	0.7933	0.0	0.9010	0.8217	0.8939	0.8812	0.0	0.6938	0.0	0.7379	0.6730	0.8837	0.8663
0.5949	1.24	160	0.4460	0.5538	0.6314	0.8813	0.9621	0.0	0.8044	0.0	0.8840	0.8692	0.9000	0.8930	0.0	0.7062	0.0	0.7445	0.6487	0.8843	0.8695
1.089	1.4	180	0.4080	0.5666	0.6331	0.8889	0.9680	0.0	0.8240	0.0	0.9200	0.8256	0.8940	0.8902	0.0	0.7323	0.0	0.7603	0.6986	0.8846	0.8762
0.959	1.55	200	0.3853	0.5675	0.6403	0.8901	0.9611	0.0	0.8645	0.0	0.8983	0.8616	0.8964	0.8975	0.0	0.7192	0.0	0.7672	0.7035	0.8848	0.8780
0.577	1.71	220	0.3735	0.5683	0.6338	0.8902	0.9760	0.0	0.7882	0.0009	0.9125	0.8542	0.9047	0.8780	0.0	0.6862	0.0009	0.8013	0.7191	0.8927	0.8769
0.9331	1.86	240	0.3605	0.5705	0.6464	0.8943	0.9648	0.0	0.8650	0.0	0.8805	0.8960	0.9181	0.9010	0.0	0.7210	0.0	0.7994	0.6841	0.8879	0.8823
0.5132	2.02	260	0.3680	0.5826	0.6573	0.8884	0.9366	0.0	0.8773	0.1041	0.9113	0.8567	0.9152	0.8694	0.0	0.6984	0.1041	0.8038	0.7009	0.9013	0.8784
0.3977	2.17	280	0.3233	0.5814	0.6498	0.9018	0.9623	0.0	0.8900	0.0	0.9162	0.8487	0.9313	0.9052	0.0	0.7370	0.0	0.8210	0.7242	0.8824	0.8891
0.6267	2.33	300	0.3394	0.5998	0.6646	0.8980	0.9735	0.0	0.8603	0.1726	0.9086	0.8262	0.9110	0.8862	0.0	0.7533	0.1726	0.8293	0.7044	0.8531	0.8871
0.7394	2.48	320	0.3483	0.5776	0.6508	0.8996	0.9566	0.0	0.9175	0.0028	0.9041	0.8487	0.9261	0.9054	0.0	0.7360	0.0028	0.8317	0.7110	0.8563	0.8874
0.4013	2.64	340	0.3393	0.5746	0.6458	0.8987	0.9734	0.0	0.8959	0.0001	0.8831	0.8420	0.9264	0.9071	0.0	0.7357	0.0001	0.8252	0.7013	0.8525	0.8859
0.3434	2.79	360	0.3745	0.6263	0.6907	0.9002	0.9516	0.0003	0.8148	0.3544	0.9583	0.8127	0.9431	0.9068	0.0003	0.7490	0.3541	0.7943	0.7221	0.8576	0.8911
0.5902	2.95	380	0.3275	0.6068	0.6696	0.9005	0.9722	0.0	0.8696	0.1949	0.9158	0.8191	0.9156	0.8941	0.0	0.7545	0.1948	0.8224	0.7303	0.8512	0.8898
0.2954	3.1	400	0.3389	0.6191	0.6862	0.9008	0.9662	0.0020	0.8410	0.3092	0.9207	0.8331	0.9316	0.9074	0.0020	0.7641	0.3056	0.8117	0.7045	0.8386	0.8913
0.4758	3.26	420	0.3122	0.6064	0.6709	0.9030	0.9682	0.0	0.9069	0.1837	0.9142	0.7957	0.9279	0.9092	0.0	0.7655	0.1834	0.8172	0.7227	0.8466	0.8923
0.6415	3.41	440	0.3109	0.6555	0.7172	0.9101	0.9728	0.0017	0.9050	0.4656	0.8885	0.8320	0.9550	0.9063	0.0017	0.7957	0.4534	0.8265	0.7512	0.8539	0.9014
0.5413	3.57	460	0.2993	0.6420	0.7045	0.9057	0.9737	0.0036	0.8782	0.4118	0.8932	0.8221	0.9487	0.9047	0.0036	0.7774	0.4103	0.8169	0.7262	0.8553	0.8967
0.2761	3.72	480	0.2802	0.6678	0.7309	0.9144	0.9619	0.0191	0.9168	0.5126	0.9190	0.8330	0.9542	0.9143	0.0191	0.7850	0.4981	0.8452	0.7427	0.8699	0.9067
0.3451	3.88	500	0.3261	0.6455	0.7056	0.9075	0.9695	0.0220	0.8345	0.4049	0.9307	0.8256	0.9520	0.9085	0.0220	0.7754	0.3995	0.8173	0.7375	0.8583	0.8988
0.2361	4.03	520	0.3118	0.6677	0.7285	0.9098	0.9683	0.0204	0.8678	0.5447	0.9117	0.8323	0.9547	0.9089	0.0204	0.7750	0.5374	0.8235	0.7509	0.8580	0.9019
0.2776	4.19	540	0.2765	0.6472	0.7121	0.9109	0.9670	0.0359	0.9025	0.3801	0.9057	0.8431	0.9507	0.9182	0.0358	0.7698	0.3764	0.8356	0.7324	0.8619	0.9029
0.3823	4.34	560	0.3514	0.6558	0.7181	0.9111	0.9592	0.0525	0.9005	0.4047	0.9183	0.8321	0.9591	0.9127	0.0523	0.7740	0.4016	0.8309	0.7587	0.8599	0.9036
0.747	4.5	580	0.3340	0.6764	0.7382	0.9130	0.9680	0.0617	0.9125	0.5379	0.8957	0.8362	0.9556	0.9146	0.0614	0.7948	0.5336	0.8279	0.7439	0.8588	0.9061
0.343	4.65	600	0.3065	0.6853	0.7474	0.9139	0.9647	0.0666	0.8925	0.5978	0.9128	0.8429	0.9542	0.9171	0.0664	0.7909	0.5886	0.8287	0.7436	0.8621	0.9074
0.2358	4.81	620	0.3229	0.6916	0.7504	0.9144	0.9719	0.0880	0.8895	0.6099	0.8976	0.8312	0.9651	0.9108	0.0874	0.7855	0.6005	0.8332	0.7563	0.8674	0.9080
0.3649	4.96	640	0.3026	0.7034	0.7644	0.9151	0.9686	0.1640	0.8804	0.6447	0.9148	0.8220	0.9562	0.9129	0.1617	0.7918	0.6134	0.8333	0.7502	0.8606	0.9101
0.201	5.12	660	0.2903	0.7089	0.7678	0.9194	0.9663	0.1156	0.9208	0.6627	0.9029	0.8379	0.9684	0.9153	0.1148	0.8061	0.6444	0.8451	0.7716	0.8648	0.9138
1.1077	5.27	680	0.2596	0.7189	0.7752	0.9221	0.9746	0.2053	0.8895	0.6365	0.9112	0.8378	0.9713	0.9192	0.2016	0.8009	0.6199	0.8479	0.7545	0.8883	0.9176
0.2368	5.43	700	0.2479	0.7271	0.7986	0.9213	0.9727	0.3222	0.8877	0.7104	0.9135	0.8259	0.9580	0.9184	0.3114	0.8003	0.5698	0.8370	0.7636	0.8894	0.9184
0.2576	5.58	720	0.3153	0.7171	0.7772	0.9173	0.9772	0.1874	0.8721	0.7060	0.9007	0.8399	0.9570	0.9152	0.1839	0.7926	0.6778	0.8408	0.7455	0.8637	0.9127
0.3873	5.74	740	0.3024	0.7395	0.7985	0.9212	0.9687	0.3518	0.9369	0.6845	0.9092	0.7793	0.9589	0.9208	0.3370	0.8203	0.6516	0.8367	0.7479	0.8622	0.9184
0.2281	5.89	760	0.2583	0.7502	0.8104	0.9227	0.9699	0.4031	0.9094	0.7032	0.9107	0.8210	0.9553	0.9229	0.3810	0.8150	0.6723	0.8431	0.7563	0.8606	0.9205
0.2202	6.05	780	0.3227	0.7612	0.8314	0.9205	0.9525	0.5297	0.9437	0.7038	0.8889	0.8332	0.9677	0.9253	0.4917	0.7900	0.6693	0.8338	0.7585	0.8599	0.9197
0.292	6.2	800	0.2558	0.7614	0.8266	0.9283	0.9710	0.4248	0.8964	0.7571	0.9185	0.8553	0.9629	0.9220	0.4076	0.8202	0.6464	0.8473	0.7915	0.8946	0.9264
0.1304	6.36	820	0.2441	0.7386	0.8067	0.9244	0.9764	0.3536	0.9007	0.6908	0.9019	0.8823	0.9412	0.9109	0.3357	0.8088	0.5524	0.8277	0.8106	0.9244	0.9221
0.1664	6.51	840	0.2565	0.7685	0.8302	0.9282	0.9728	0.4257	0.8564	0.7880	0.9277	0.8793	0.9614	0.9244	0.3944	0.8006	0.7192	0.8439	0.7941	0.9032	0.9263
0.4424	6.67	860	0.2667	0.7808	0.8611	0.9332	0.9680	0.5948	0.9061	0.7883	0.9039	0.8938	0.9727	0.9362	0.5495	0.8165	0.6008	0.8453	0.8086	0.9083	0.9326
0.3753	6.82	880	0.2585	0.7353	0.8117	0.9236	0.9765	0.4103	0.8985	0.7286	0.9126	0.7922	0.9633	0.9221	0.3532	0.8025	0.5636	0.8374	0.7631	0.9054	0.9215
0.2642	6.98	900	0.2955	0.7734	0.8443	0.9247	0.9586	0.5628	0.9093	0.7641	0.9085	0.8357	0.9708	0.9240	0.5068	0.8133	0.6949	0.8429	0.7634	0.8685	0.9238
0.5008	7.13	920	0.3048	0.7703	0.8430	0.9262	0.9712	0.5688	0.8871	0.7660	0.9077	0.8329	0.9676	0.9260	0.5015	0.8170	0.6626	0.8502	0.7697	0.8651	0.9251
0.169	7.29	940	0.2969	0.7778	0.8528	0.9252	0.9644	0.6690	0.9288	0.7274	0.8966	0.8274	0.9559	0.9263	0.5909	0.8081	0.6410	0.8444	0.7733	0.8608	0.9246
0.2429	7.44	960	0.3385	0.7906	0.8551	0.9283	0.9626	0.6420	0.9047	0.7576	0.9131	0.8281	0.9776	0.9322	0.5836	0.8160	0.7156	0.8415	0.7752	0.8701	0.9276
0.1542	7.6	980	0.3259	0.7882	0.8574	0.9264	0.9646	0.6734	0.9031	0.7509	0.9007	0.8403	0.9686	0.9298	0.5949	0.8138	0.7056	0.8366	0.7707	0.8663	0.9258
0.1378	7.75	1000	0.3257	0.7815	0.8589	0.9249	0.9675	0.6849	0.8784	0.7864	0.9151	0.8219	0.9580	0.9265	0.5842	0.8076	0.6769	0.8375	0.7767	0.8610	0.9242
0.2144	7.91	1020	0.3347	0.7889	0.8561	0.9276	0.9578	0.6604	0.9128	0.7455	0.9176	0.8261	0.9729	0.9317	0.5932	0.8189	0.7046	0.8416	0.7636	0.8688	0.9270
0.1011	8.06	1040	0.3099	0.7831	0.8546	0.9267	0.9655	0.6394	0.8900	0.7685	0.9091	0.8402	0.9695	0.9289	0.5594	0.8137	0.6999	0.8434	0.7682	0.8685	0.9260
0.1543	8.22	1060	0.3291	0.7921	0.8667	0.9294	0.9667	0.7434	0.9074	0.7434	0.9025	0.8251	0.9781	0.9357	0.6127	0.8145	0.6936	0.8460	0.7703	0.8720	0.9289
0.2624	8.37	1080	0.3284	0.7884	0.8593	0.9247	0.9644	0.6802	0.8663	0.7915	0.9233	0.8317	0.9580	0.9297	0.5924	0.8031	0.7343	0.8332	0.7633	0.8631	0.9241
0.1817	8.53	1100	0.3187	0.7848	0.8557	0.9255	0.9705	0.6361	0.9041	0.7843	0.8918	0.8460	0.9573	0.9265	0.5584	0.8179	0.7210	0.8371	0.7698	0.8626	0.9248
0.1702	8.68	1120	0.3412	0.7737	0.8545	0.9252	0.9680	0.6271	0.8745	0.8030	0.9108	0.8296	0.9682	0.9285	0.5294	0.8047	0.6748	0.8430	0.7686	0.8673	0.9245

Framework versions

Transformers 4.37.0
Pytorch 2.1.2
Datasets 2.18.0
Tokenizers 0.15.1

Task: Image Segmentation
Library: transformers
Type: segformer
License: other
Created: Mar 6, 2024
Updated: Mar 6, 2024