This model is a fine-tuned version of nvidia/mit-b0 on the segments/sidewalk-semantic dataset. It achieves the following results on the evaluation set:

Loss: 1.3483
Mean Iou: 0.1474
Mean Accuracy: 0.1958
Overall Accuracy: 0.7227
Accuracy Unlabeled: nan
Accuracy Flat-road: 0.7892
Accuracy Flat-sidewalk: 0.9075
Accuracy Flat-crosswalk: 0.0
Accuracy Flat-cyclinglane: 0.1029
Accuracy Flat-parkingdriveway: 0.0001
Accuracy Flat-railtrack: nan
Accuracy Flat-curb: 0.0
Accuracy Human-person: 0.0
Accuracy Human-rider: 0.0
Accuracy Vehicle-car: 0.8749
Accuracy Vehicle-truck: 0.0
Accuracy Vehicle-bus: 0.0
Accuracy Vehicle-tramtrain: nan
Accuracy Vehicle-motorcycle: 0.0
Accuracy Vehicle-bicycle: 0.0
Accuracy Vehicle-caravan: 0.0
Accuracy Vehicle-cartrailer: 0.0
Accuracy Construction-building: 0.8943
Accuracy Construction-door: 0.0
Accuracy Construction-wall: 0.0
Accuracy Construction-fenceguardrail: 0.0
Accuracy Construction-bridge: 0.0
Accuracy Construction-tunnel: nan
Accuracy Construction-stairs: 0.0
Accuracy Object-pole: 0.0
Accuracy Object-trafficsign: 0.0
Accuracy Object-trafficlight: 0.0
Accuracy Nature-vegetation: 0.9118
Accuracy Nature-terrain: 0.6546
Accuracy Sky: 0.9352
Accuracy Void-ground: 0.0
Accuracy Void-dynamic: 0.0
Accuracy Void-static: 0.0
Accuracy Void-unclear: 0.0
Iou Unlabeled: nan
Iou Flat-road: 0.4282
Iou Flat-sidewalk: 0.7768
Iou Flat-crosswalk: 0.0
Iou Flat-cyclinglane: 0.1021
Iou Flat-parkingdriveway: 0.0001
Iou Flat-railtrack: nan
Iou Flat-curb: 0.0
Iou Human-person: 0.0
Iou Human-rider: 0.0
Iou Vehicle-car: 0.6372
Iou Vehicle-truck: 0.0
Iou Vehicle-bus: 0.0
Iou Vehicle-tramtrain: nan
Iou Vehicle-motorcycle: 0.0
Iou Vehicle-bicycle: 0.0
Iou Vehicle-caravan: 0.0
Iou Vehicle-cartrailer: 0.0
Iou Construction-building: 0.5530
Iou Construction-door: 0.0
Iou Construction-wall: 0.0
Iou Construction-fenceguardrail: 0.0
Iou Construction-bridge: 0.0
Iou Construction-tunnel: nan
Iou Construction-stairs: 0.0
Iou Object-pole: 0.0
Iou Object-trafficsign: 0.0
Iou Object-trafficlight: 0.0
Iou Nature-vegetation: 0.7392
Iou Nature-terrain: 0.5009
Iou Sky: 0.8328
Iou Void-ground: 0.0
Iou Void-dynamic: 0.0
Iou Void-static: 0.0
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: 2
eval_batch_size: 2
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Flat-road	Accuracy Flat-sidewalk	Accuracy Flat-cyclinglane	Accuracy Flat-parkingdriveway	Accuracy Flat-railtrack	Accuracy Flat-curb	Accuracy Human-person	Accuracy Vehicle-car	Accuracy Vehicle-tramtrain	Accuracy Construction-building	Accuracy Construction-wall	Accuracy Construction-fenceguardrail	Accuracy Construction-tunnel	Accuracy Construction-stairs	Accuracy Object-pole	Accuracy Nature-vegetation	Accuracy Nature-terrain	Accuracy Sky	Accuracy Void-ground	Accuracy Void-static	Iou Unlabeled	Iou Flat-road	Iou Flat-sidewalk	Iou Flat-cyclinglane	Iou Flat-parkingdriveway	Iou Flat-railtrack	Iou Flat-curb	Iou Human-person	Iou Vehicle-car	Iou Vehicle-tramtrain	Iou Construction-building	Iou Construction-wall	Iou Construction-fenceguardrail	Iou Construction-tunnel	Iou Construction-stairs	Iou Object-pole	Iou Nature-vegetation	Iou Nature-terrain	Iou Sky	Iou Void-ground	Iou Void-static
2.8103	0.06	25	3.0462	0.0835	0.1351	0.5790	nan	0.1892	0.9330	0.0048	0.0006	nan	0.0005	0.0002	0.5793	nan	0.8849	0.0000	0.0002	nan	0.0003	0.0009	0.6599	0.4398	0.4779	0.0004	0.0151	0.0	0.1368	0.6091	0.0046	0.0006	0.0	0.0005	0.0002	0.4916	0.0	0.3537	0.0000	0.0002	0.0	0.0003	0.0008	0.5851	0.2698	0.4572	0.0004	0.0123
2.3833	0.12	50	2.3708	0.1025	0.1539	0.6295	nan	0.6185	0.8261	0.0007	0.0003	nan	0.0000	0.0	0.7749	nan	0.9283	0.0	0.0	nan	0.0	0.0	0.8688	0.0568	0.6944	0.0	0.0016	nan	0.3347	0.6672	0.0007	0.0003	nan	0.0000	0.0	0.5428	0.0	0.4369	0.0	0.0	nan	0.0	0.0	0.6274	0.0451	0.6247	0.0	0.0015
2.1946	0.19	75	1.9680	0.1145	0.1612	0.6615	nan	0.6725	0.8726	0.0011	0.0000	nan	0.0	0.0	0.7496	nan	0.9138	0.0	0.0	nan	0.0	0.0	0.9185	0.0333	0.8355	0.0	0.0000	nan	0.3738	0.7037	0.0011	0.0000	nan	0.0	0.0	0.5385	nan	0.4836	0.0	0.0	nan	0.0	0.0	0.6493	0.0311	0.7684	0.0	0.0000
1.959	0.25	100	1.8828	0.1179	0.1636	0.6727	nan	0.6801	0.8928	0.0006	0.0001	nan	0.0	0.0	0.7736	nan	0.8784	0.0	0.0	nan	0.0	0.0	0.9482	0.0389	0.8584	0.0	0.0000	nan	0.3884	0.7151	0.0006	0.0001	nan	0.0	0.0	0.5705	nan	0.5160	0.0	0.0	nan	0.0	0.0	0.6418	0.0360	0.7878	0.0	0.0000
1.8759	0.31	125	1.7092	0.1260	0.1758	0.6862	nan	0.7247	0.8976	0.0000	0.0000	nan	0.0	0.0	0.9094	nan	0.8479	0.0	0.0	nan	0.0	0.0	0.9024	0.2625	0.9054	0.0	0.0	nan	0.3977	0.7307	0.0000	0.0000	nan	0.0	0.0	0.5237	nan	0.5329	0.0	0.0	nan	0.0	0.0	0.6911	0.2245	0.8065	0.0	0.0
2.0333	0.38	150	1.5558	0.1267	0.1751	0.6898	nan	0.7565	0.8952	0.0007	0.0000	nan	0.0	0.0	0.8792	nan	0.8874	0.0	0.0	nan	0.0	0.0	0.9104	0.1928	0.9055	0.0	0.0	nan	0.4083	0.7455	0.0007	0.0000	nan	0.0	0.0	0.5928	nan	0.5267	0.0	0.0	nan	0.0	0.0	0.6820	0.1637	0.8069	0.0	0.0
1.8985	0.44	175	1.5370	0.1277	0.1752	0.6939	nan	0.7438	0.9025	0.0068	0.0000	nan	0.0	0.0	0.8237	nan	0.8984	0.0	0.0	nan	0.0	0.0	0.9307	0.1976	0.9266	0.0	0.0	nan	0.4223	0.7553	0.0068	0.0000	nan	0.0	0.0	0.6077	nan	0.5131	0.0	0.0	nan	0.0	0.0	0.6724	0.1691	0.8133	0.0	0.0
1.7908	0.5	200	1.4854	0.1339	0.1843	0.7020	nan	0.7612	0.9068	0.0012	0.0	nan	0.0	0.0	0.9215	nan	0.8381	0.0	0.0	nan	0.0	0.0	0.9078	0.4480	0.9274	0.0	0.0	nan	0.4095	0.7591	0.0012	0.0	nan	0.0	0.0	0.5471	nan	0.5386	0.0	0.0	nan	0.0	0.0	0.7168	0.3673	0.8119	0.0	0.0
1.4371	0.56	225	1.4176	0.1367	0.1830	0.7079	nan	0.7035	0.9303	0.0255	0.0001	nan	0.0	0.0	0.8694	nan	0.8848	0.0	0.0	nan	0.0	0.0	0.9287	0.4230	0.9065	0.0	0.0	nan	0.4233	0.7534	0.0255	0.0001	nan	0.0	0.0	0.6246	nan	0.5331	0.0	0.0	nan	0.0	0.0	0.7049	0.3427	0.8312	0.0	0.0
1.4506	0.62	250	1.4011	0.1350	0.1827	0.7079	nan	0.6936	0.9362	0.0364	0.0000	nan	0.0	0.0	0.8955	nan	0.8612	0.0	0.0	nan	0.0	0.0	0.9299	0.3720	0.9394	0.0	0.0	nan	0.4251	0.7533	0.0363	0.0000	nan	0.0	0.0	0.5961	nan	0.5472	0.0	0.0	nan	0.0	0.0	0.6990	0.3026	0.8250	0.0	0.0
1.3095	0.69	275	1.4039	0.1398	0.1861	0.7134	nan	0.6873	0.9430	0.0260	0.0000	nan	0.0	0.0	0.8807	nan	0.8960	0.0	0.0	nan	0.0	0.0	0.9156	0.5055	0.9137	0.0	0.0	nan	0.4187	0.7531	0.0260	0.0000	nan	0.0	0.0	0.6172	nan	0.5457	0.0	0.0	nan	0.0	0.0	0.7261	0.4143	0.8343	0.0	0.0
1.4218	0.75	300	1.3735	0.1370	0.1856	0.7082	nan	0.7701	0.8980	0.0524	0.0003	nan	0.0	0.0	0.8776	nan	0.8762	0.0	0.0	nan	0.0	0.0	0.9444	0.4052	0.9279	0.0	0.0	nan	0.4267	0.7668	0.0522	0.0003	nan	0.0	0.0	0.6148	nan	0.5481	0.0	0.0	nan	0.0	0.0	0.6931	0.3263	0.8172	0.0	0.0
1.3986	0.81	325	1.3224	0.1451	0.1917	0.7223	nan	0.7354	0.9296	0.1261	0.0002	nan	0.0	0.0	0.8747	nan	0.8946	0.0	0.0	nan	0.0	0.0	0.9269	0.5292	0.9261	0.0	0.0	nan	0.4389	0.7700	0.1253	0.0002	nan	0.0	0.0	0.6395	nan	0.5541	0.0	0.0	nan	0.0	0.0	0.7200	0.4083	0.8409	0.0	0.0
1.3335	0.88	350	1.2909	0.1454	0.1921	0.7230	nan	0.7157	0.9388	0.1024	0.0003	nan	0.0	0.0	0.8896	nan	0.8932	0.0	0.0	nan	0.0	0.0	0.9207	0.5667	0.9276	0.0	0.0	nan	0.4347	0.7657	0.1020	0.0003	nan	0.0	0.0	0.6232	nan	0.5533	0.0	0.0	nan	0.0	0.0	0.7300	0.4607	0.8376	0.0	0.0
1.6376	0.94	375	1.3109	0.1476	0.1964	0.7253	nan	0.7617	0.9215	0.1225	0.0002	nan	0.0	0.0	0.9008	nan	0.8842	0.0	0.0	nan	0.0	0.0	0.9118	0.6529	0.9335	0.0	0.0	nan	0.4357	0.7750	0.1216	0.0002	nan	0.0	0.0	0.6139	nan	0.5572	0.0	0.0	nan	0.0	0.0	0.7387	0.5011	0.8333	0.0	0.0
1.4354	1.0	400	1.3483	0.1474	0.1958	0.7227	nan	0.7892	0.9075	0.1029	0.0001	nan	0.0	0.0	0.8749	nan	0.8943	0.0	0.0	nan	0.0	0.0	0.9118	0.6546	0.9352	0.0	0.0	nan	0.4282	0.7768	0.1021	0.0001	nan	0.0	0.0	0.6372	nan	0.5530	0.0	0.0	nan	0.0	0.0	0.7392	0.5009	0.8328	0.0	0.0

Framework versions

Transformers 4.35.2
Pytorch 2.1.0+cu118
Datasets 2.15.0
Tokenizers 0.15.0

piecurus/segformer-finetuned-segments-opit

segformer-finetuned-segments-opit

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

piecurus/
segformer-finetuned-segments-opit