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

Loss: 0.8650
Mean Iou: 0.1979
Mean Accuracy: 0.2600
Overall Accuracy: 0.7844
Accuracy Unlabeled: nan
Accuracy Flat-road: 0.7636
Accuracy Flat-sidewalk: 0.9369
Accuracy Flat-crosswalk: 0.6233
Accuracy Flat-cyclinglane: 0.5039
Accuracy Flat-parkingdriveway: 0.3350
Accuracy Flat-railtrack: nan
Accuracy Flat-curb: 0.3609
Accuracy Human-person: 0.0041
Accuracy Human-rider: 0.0
Accuracy Vehicle-car: 0.9008
Accuracy Vehicle-truck: 0.0
Accuracy Vehicle-bus: 0.0
Accuracy Vehicle-tramtrain: 0.0
Accuracy Vehicle-motorcycle: 0.0
Accuracy Vehicle-bicycle: 0.0005
Accuracy Vehicle-caravan: 0.0
Accuracy Vehicle-cartrailer: 0.0
Accuracy Construction-building: 0.8702
Accuracy Construction-door: 0.0
Accuracy Construction-wall: 0.3334
Accuracy Construction-fenceguardrail: 0.1194
Accuracy Construction-bridge: 0.0
Accuracy Construction-tunnel: 0.0
Accuracy Construction-stairs: 0.0
Accuracy Object-pole: 0.0637
Accuracy Object-trafficsign: 0.0
Accuracy Object-trafficlight: 0.0
Accuracy Nature-vegetation: 0.9057
Accuracy Nature-terrain: 0.8349
Accuracy Sky: 0.9210
Accuracy Void-ground: 0.0339
Accuracy Void-dynamic: 0.0032
Accuracy Void-static: 0.0657
Accuracy Void-unclear: 0.0
Iou Unlabeled: 0.0
Iou Flat-road: 0.6053
Iou Flat-sidewalk: 0.7972
Iou Flat-crosswalk: 0.4947
Iou Flat-cyclinglane: 0.4639
Iou Flat-parkingdriveway: 0.2544
Iou Flat-railtrack: 0.0
Iou Flat-curb: 0.2330
Iou Human-person: 0.0041
Iou Human-rider: 0.0
Iou Vehicle-car: 0.6748
Iou Vehicle-truck: 0.0
Iou Vehicle-bus: 0.0
Iou Vehicle-tramtrain: 0.0
Iou Vehicle-motorcycle: 0.0
Iou Vehicle-bicycle: 0.0005
Iou Vehicle-caravan: 0.0
Iou Vehicle-cartrailer: 0.0
Iou Construction-building: 0.6517
Iou Construction-door: 0.0
Iou Construction-wall: 0.2393
Iou Construction-fenceguardrail: 0.0990
Iou Construction-bridge: 0.0
Iou Construction-tunnel: 0.0
Iou Construction-stairs: 0.0
Iou Object-pole: 0.0522
Iou Object-trafficsign: 0.0
Iou Object-trafficlight: 0.0
Iou Nature-vegetation: 0.7891
Iou Nature-terrain: 0.6367
Iou Sky: 0.8417
Iou Void-ground: 0.0306
Iou Void-dynamic: 0.0031
Iou Void-static: 0.0567
Iou Void-unclear: 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: 6e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Flat-road	Accuracy Flat-sidewalk	Accuracy Flat-crosswalk	Accuracy Flat-cyclinglane	Accuracy Flat-parkingdriveway	Accuracy Flat-railtrack	Accuracy Flat-curb	Accuracy Human-person	Accuracy Human-rider	Accuracy Vehicle-car	Accuracy Vehicle-bus	Accuracy Vehicle-motorcycle	Accuracy Vehicle-bicycle	Accuracy Vehicle-cartrailer	Accuracy Construction-building	Accuracy Construction-door	Accuracy Construction-wall	Accuracy Construction-fenceguardrail	Accuracy Object-pole	Accuracy Nature-vegetation	Accuracy Nature-terrain	Accuracy Sky	Accuracy Void-ground	Accuracy Void-dynamic	Accuracy Void-static	Iou Flat-road	Iou Flat-sidewalk	Iou Flat-crosswalk	Iou Flat-cyclinglane	Iou Flat-parkingdriveway	Iou Flat-curb	Iou Human-person	Iou Human-rider	Iou Vehicle-car	Iou Vehicle-bus	Iou Vehicle-motorcycle	Iou Vehicle-bicycle	Iou Vehicle-cartrailer	Iou Construction-building	Iou Construction-door	Iou Construction-wall	Iou Construction-fenceguardrail	Iou Object-pole	Iou Nature-vegetation	Iou Nature-terrain	Iou Sky	Iou Void-ground	Iou Void-dynamic	Iou Void-static
2.5085	0.2	20	2.7797	0.0732	0.1076	0.4652	nan	0.5841	0.7791	0.0055	0.2836	0.0034	nan	0.0	0.0	0.0	0.5620	0.0198	0.0004	0.0004	0.0535	0.0891	0.0012	0.0003	0.0010	0.0	0.3935	0.6589	0.1063	0.0	0.0081	0.0	0.3763	0.6424	0.0019	0.2150	0.0033	0.0	0.0	0.0	0.2835	0.0001	0.0001	0.0001	0.0000	0.0859	0.0011	0.0002	0.0010	0.0	0.3843	0.4577	0.1061	0.0	0.0024	0.0
2.4875	0.4	40	1.7587	0.1295	0.1722	0.6580	nan	0.7729	0.8313	0.0254	0.3069	0.0016	nan	0.0005	0.0	0.0	0.7856	0.0017	0.0	0.0	0.0044	0.6115	0.0009	0.0148	0.0122	0.0032	0.8782	0.6807	0.7454	0.0	0.0067	0.0	0.4677	0.7337	0.0129	0.2895	0.0016	0.0005	0.0	0.0	0.5336	0.0000	0.0	0.0	0.0000	0.4970	0.0009	0.0130	0.0114	0.0032	0.7164	0.5249	0.7231	0.0	0.0030	0.0
1.4979	0.6	60	1.3253	0.1478	0.2057	0.6899	nan	0.8068	0.8066	0.1982	0.4657	0.2635	nan	0.0285	0.0	0.0	0.8709	0.0017	0.0	0.0005	0.0	0.6424	0.0000	0.1611	0.0051	0.0138	0.8751	0.7601	0.8505	0.0261	0.0108	0.0001	0.4953	0.7315	0.1383	0.3556	0.1491	0.0266	0.0	0.0	0.4975	0.0001	0.0	0.0005	0.0	0.5333	0.0000	0.1061	0.0049	0.0122	0.7534	0.5484	0.7903	0.0243	0.0045	0.0001
1.2341	0.8	80	1.2097	0.1612	0.2131	0.7259	nan	0.5034	0.9552	0.5156	0.4170	0.1216	nan	0.1390	0.0	0.0003	0.8090	0.0	0.0	0.0	0.0	0.8904	0.0	0.1239	0.0350	0.0102	0.8813	0.7535	0.8759	0.0	0.0013	0.0003	0.4375	0.7135	0.3661	0.3521	0.1009	0.1176	0.0	0.0000	0.6389	0.0	0.0	0.0	0.0	0.5992	0.0	0.0962	0.0306	0.0092	0.7711	0.5926	0.8134	0.0	0.0011	0.0003
1.594	1.0	100	1.0384	0.1687	0.2260	0.7523	nan	0.8702	0.8883	0.3212	0.4611	0.1767	nan	0.1078	0.0	0.0013	0.8716	0.0	0.0	0.0023	0.0	0.8338	0.0	0.2006	0.0366	0.0374	0.8996	0.8280	0.8915	0.0	0.0008	0.0306	0.5099	0.7920	0.2474	0.4321	0.1507	0.0913	0.0	0.0003	0.6235	0.0	0.0	0.0020	0.0	0.6274	0.0	0.1603	0.0337	0.0300	0.7659	0.5944	0.8171	0.0	0.0008	0.0260
0.7624	1.2	120	0.9402	0.1852	0.2479	0.7620	nan	0.6183	0.9445	0.6022	0.5182	0.2131	nan	0.3901	0.0	0.0008	0.8980	0.0	0.0	0.0001	0.0	0.8681	0.0000	0.2897	0.1158	0.0470	0.8860	0.8323	0.9143	0.0032	0.0017	0.0376	0.5142	0.7737	0.4531	0.4391	0.1737	0.2176	0.0	0.0002	0.6419	0.0	0.0	0.0001	0.0	0.6411	0.0000	0.2041	0.0941	0.0376	0.7853	0.6371	0.8329	0.0029	0.0016	0.0328
1.1404	1.4	140	0.9037	0.1937	0.2559	0.7761	nan	0.7507	0.9337	0.6121	0.5261	0.2748	nan	0.2728	0.0007	0.0	0.8756	0.0	0.0	0.0002	0.0	0.8461	0.0000	0.4180	0.1162	0.0773	0.9055	0.8279	0.9024	0.0404	0.0019	0.0623	0.5682	0.7933	0.4733	0.4630	0.2328	0.1898	0.0007	0.0	0.6841	0.0	0.0	0.0002	0.0	0.6414	0.0000	0.2357	0.0930	0.0561	0.7862	0.6368	0.8350	0.0348	0.0018	0.0532
0.7949	1.6	160	0.8834	0.1959	0.2559	0.7833	nan	0.7455	0.9427	0.6186	0.5220	0.3557	nan	0.3018	0.0044	0.0	0.9011	0.0	0.0	0.0012	0.0	0.8695	0.0	0.3135	0.1094	0.0602	0.9121	0.7871	0.9168	0.0254	0.0012	0.0554	0.5954	0.7952	0.4864	0.4753	0.2686	0.2131	0.0044	0.0	0.6642	0.0	0.0	0.0011	0.0	0.6427	0.0	0.2290	0.0919	0.0486	0.7904	0.6367	0.8404	0.0234	0.0011	0.0493
1.1158	1.8	180	0.8739	0.1967	0.2579	0.7829	nan	0.7531	0.9429	0.6044	0.4876	0.3491	nan	0.3338	0.0041	0.0	0.8976	0.0	0.0	0.0005	0.0	0.8643	0.0	0.3477	0.1201	0.0602	0.9064	0.8280	0.9161	0.0301	0.0027	0.0612	0.6023	0.7929	0.4754	0.4542	0.2581	0.2261	0.0041	0.0	0.6759	0.0	0.0	0.0005	0.0	0.6505	0.0	0.2436	0.0991	0.0493	0.7883	0.6363	0.8424	0.0275	0.0026	0.0538
0.9007	2.0	200	0.8650	0.1979	0.2600	0.7844	nan	0.7636	0.9369	0.6233	0.5039	0.3350	nan	0.3609	0.0041	0.0	0.9008	0.0	0.0	0.0005	0.0	0.8702	0.0	0.3334	0.1194	0.0637	0.9057	0.8349	0.9210	0.0339	0.0032	0.0657	0.6053	0.7972	0.4947	0.4639	0.2544	0.2330	0.0041	0.0	0.6748	0.0	0.0	0.0005	0.0	0.6517	0.0	0.2393	0.0990	0.0522	0.7891	0.6367	0.8417	0.0306	0.0031	0.0567

Framework versions

Transformers 4.48.0
Pytorch 2.1.1+cu118
Datasets 3.2.0
Tokenizers 0.21.0

Kohei3/my-fine-tuned-model

my-fine-tuned-model

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Kohei3/
my-fine-tuned-model