PieroPR/mit-b0-finetuned-sidewalks

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

• Train Loss: 0.5658
• Validation Loss: 0.6820
• Validation Mean Iou: 0.2736
• Validation Mean Accuracy: 0.3393
• Validation Overall Accuracy: 0.8140
• Validation Accuracy Unlabeled: 0.0
• Validation Accuracy Flat-road: 0.8209
• Validation Accuracy Flat-sidewalk: 0.9071
• Validation Accuracy Flat-crosswalk: 0.8663
• Validation Accuracy Flat-cyclinglane: 0.7009
• Validation Accuracy Flat-parkingdriveway: 0.4370
• Validation Accuracy Flat-railtrack: nan
• Validation Accuracy Flat-curb: 0.4513
• Validation Accuracy Human-person: 0.7244
• Validation Accuracy Human-rider: 0.0
• Validation Accuracy Vehicle-car: 0.9263
• Validation Accuracy Vehicle-truck: 0.0
• Validation Accuracy Vehicle-bus: 0.0
• Validation Accuracy Vehicle-tramtrain: 0.0
• Validation Accuracy Vehicle-motorcycle: 0.0
• Validation Accuracy Vehicle-bicycle: 0.3297
• Validation Accuracy Vehicle-caravan: 0.0
• Validation Accuracy Vehicle-cartrailer: nan
• Validation Accuracy Construction-building: 0.8640
• Validation Accuracy Construction-door: 0.0239
• Validation Accuracy Construction-wall: 0.4192
• Validation Accuracy Construction-fenceguardrail: 0.3167
• Validation Accuracy Construction-bridge: 0.0
• Validation Accuracy Construction-tunnel: 0.0
• Validation Accuracy Construction-stairs: 0.0
• Validation Accuracy Object-pole: 0.3078
• Validation Accuracy Object-trafficsign: 0.0000
• Validation Accuracy Object-trafficlight: 0.0
• Validation Accuracy Nature-vegetation: 0.9553
• Validation Accuracy Nature-terrain: 0.8359
• Validation Accuracy Sky: 0.9632
• Validation Accuracy Void-ground: 0.0001
• Validation Accuracy Void-dynamic: 0.0059
• Validation Accuracy Void-static: 0.3424
• Validation Accuracy Void-unclear: 0.0
• Validation Iou Unlabeled: 0.0
• Validation Iou Flat-road: 0.6646
• Validation Iou Flat-sidewalk: 0.8269
• Validation Iou Flat-crosswalk: 0.5585
• Validation Iou Flat-cyclinglane: 0.6586
• Validation Iou Flat-parkingdriveway: 0.2672
• Validation Iou Flat-railtrack: nan
• Validation Iou Flat-curb: 0.3489
• Validation Iou Human-person: 0.5512
• Validation Iou Human-rider: 0.0
• Validation Iou Vehicle-car: 0.7427
• Validation Iou Vehicle-truck: 0.0
• Validation Iou Vehicle-bus: 0.0
• Validation Iou Vehicle-tramtrain: 0.0
• Validation Iou Vehicle-motorcycle: 0.0
• Validation Iou Vehicle-bicycle: 0.2798
• Validation Iou Vehicle-caravan: 0.0
• Validation Iou Vehicle-cartrailer: nan
• Validation Iou Construction-building: 0.7027
• Validation Iou Construction-door: 0.0239
• Validation Iou Construction-wall: 0.2868
• Validation Iou Construction-fenceguardrail: 0.2483
• Validation Iou Construction-bridge: 0.0
• Validation Iou Construction-tunnel: 0.0
• Validation Iou Construction-stairs: 0.0
• Validation Iou Object-pole: 0.2342
• Validation Iou Object-trafficsign: 0.0000
• Validation Iou Object-trafficlight: 0.0
• Validation Iou Nature-vegetation: 0.8203
• Validation Iou Nature-terrain: 0.6535
• Validation Iou Sky: 0.9096
• Validation Iou Void-ground: 0.0001
• Validation Iou Void-dynamic: 0.0041
• Validation Iou Void-static: 0.2463
• Validation Iou Void-unclear: 0.0
• Epoch: 4

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:

• optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': 6e-05, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Validation Loss	Validation Mean Iou	Validation Mean Accuracy	Validation Overall Accuracy	Validation Accuracy Unlabeled	Validation Accuracy Flat-road	Validation Accuracy Flat-sidewalk	Validation Accuracy Flat-crosswalk	Validation Accuracy Flat-cyclinglane	Validation Accuracy Flat-parkingdriveway	Validation Accuracy Flat-railtrack	Validation Accuracy Flat-curb	Validation Accuracy Human-person	Validation Accuracy Human-rider	Validation Accuracy Vehicle-car	Validation Accuracy Vehicle-truck	Validation Accuracy Vehicle-bus	Validation Accuracy Vehicle-tramtrain	Validation Accuracy Vehicle-motorcycle	Validation Accuracy Vehicle-bicycle	Validation Accuracy Vehicle-caravan	Validation Accuracy Vehicle-cartrailer	Validation Accuracy Construction-building	Validation Accuracy Construction-door	Validation Accuracy Construction-wall	Validation Accuracy Construction-fenceguardrail	Validation Accuracy Construction-bridge	Validation Accuracy Construction-tunnel	Validation Accuracy Construction-stairs	Validation Accuracy Object-pole	Validation Accuracy Object-trafficsign	Validation Accuracy Object-trafficlight	Validation Accuracy Nature-vegetation	Validation Accuracy Nature-terrain	Validation Accuracy Sky	Validation Accuracy Void-ground	Validation Accuracy Void-dynamic	Validation Accuracy Void-static	Validation Accuracy Void-unclear	Validation Iou Unlabeled	Validation Iou Flat-road	Validation Iou Flat-sidewalk	Validation Iou Flat-crosswalk	Validation Iou Flat-cyclinglane	Validation Iou Flat-parkingdriveway	Validation Iou Flat-railtrack	Validation Iou Flat-curb	Validation Iou Human-person	Validation Iou Human-rider	Validation Iou Vehicle-car	Validation Iou Vehicle-truck	Validation Iou Vehicle-bus	Validation Iou Vehicle-tramtrain	Validation Iou Vehicle-motorcycle	Validation Iou Vehicle-bicycle	Validation Iou Vehicle-caravan	Validation Iou Vehicle-cartrailer	Validation Iou Construction-building	Validation Iou Construction-door	Validation Iou Construction-wall	Validation Iou Construction-fenceguardrail	Validation Iou Construction-bridge	Validation Iou Construction-tunnel	Validation Iou Construction-stairs	Validation Iou Object-pole	Validation Iou Object-trafficsign	Validation Iou Object-trafficlight	Validation Iou Nature-vegetation	Validation Iou Nature-terrain	Validation Iou Sky	Validation Iou Void-ground	Validation Iou Void-dynamic	Validation Iou Void-static	Validation Iou Void-unclear	Epoch
1.4417	0.9458	0.1720	0.2164	0.7492	0.0	0.7213	0.9392	0.0032	0.6615	0.0978	nan	0.0859	0.0166	0.0	0.8492	0.0	0.0	0.0	0.0	0.0004	0.0	nan	0.8935	0.0	0.2675	0.0126	0.0	0.0	0.0	0.0139	0.0	0.0	0.9325	0.7108	0.9198	0.0	0.0	0.0155	0.0	0.0	0.5252	0.7670	0.0032	0.5555	0.0878	nan	0.0719	0.0165	0.0	0.6335	0.0	0.0	0.0	0.0	0.0004	0.0	nan	0.6036	0.0	0.1696	0.0124	0.0	0.0	0.0	0.0136	0.0	0.0	0.7625	0.5888	0.8511	0.0	0.0	0.0148	0.0	0
0.8432	0.7725	0.2355	0.2899	0.7998	0.0	0.8920	0.9053	0.6734	0.7470	0.3021	nan	0.3411	0.3125	0.0	0.9194	0.0	0.0	0.0	0.0	0.2151	0.0	nan	0.8866	0.0	0.3137	0.1988	0.0	0.0	0.0	0.1288	0.0	0.0	0.9490	0.7460	0.9398	0.0	0.0	0.0970	0.0	0.0	0.6620	0.8339	0.5670	0.5954	0.2320	nan	0.2613	0.2838	0.0	0.6686	0.0	0.0	0.0	0.0	0.1247	0.0	nan	0.6523	0.0	0.2020	0.1600	0.0	0.0	0.0	0.1165	0.0	0.0	0.7997	0.6352	0.8898	0.0	0.0	0.0860	0.0	1
0.7035	0.7581	0.2508	0.3187	0.7947	0.0	0.8945	0.8435	0.8598	0.7456	0.4480	nan	0.3224	0.5241	0.0	0.9059	0.0	0.0	0.0	0.0	0.3407	0.0	nan	0.9156	0.0	0.2807	0.2499	0.0	0.0	0.0	0.2601	0.0	0.0	0.9390	0.8018	0.9606	0.0	0.0049	0.2197	0.0	0.0	0.6317	0.8090	0.4994	0.6327	0.2302	nan	0.2613	0.4452	0.0	0.6982	0.0	0.0	0.0	0.0	0.1719	0.0	nan	0.6719	0.0	0.2417	0.1960	0.0	0.0	0.0	0.2143	0.0	0.0	0.8222	0.6768	0.8946	0.0	0.0033	0.1775	0.0	2
0.6112	0.6224	0.2727	0.3304	0.8278	0.0	0.8344	0.9527	0.8526	0.7185	0.3794	nan	0.4402	0.6559	0.0	0.9143	0.0	0.0	0.0	0.0	0.3014	0.0	nan	0.9239	0.0	0.4043	0.3335	0.0	0.0	0.0	0.2741	0.0	0.0	0.9102	0.8604	0.9556	0.0	0.0062	0.1850	0.0	0.0	0.7346	0.8435	0.7084	0.6373	0.2833	nan	0.3242	0.5212	0.0	0.7058	0.0	0.0	0.0	0.0	0.2147	0.0	nan	0.7120	0.0	0.2865	0.2758	0.0	0.0	0.0	0.2228	0.0	0.0	0.8232	0.6333	0.9095	0.0	0.0054	0.1570	0.0	3
0.5658	0.6820	0.2736	0.3393	0.8140	0.0	0.8209	0.9071	0.8663	0.7009	0.4370	nan	0.4513	0.7244	0.0	0.9263	0.0	0.0	0.0	0.0	0.3297	0.0	nan	0.8640	0.0239	0.4192	0.3167	0.0	0.0	0.0	0.3078	0.0000	0.0	0.9553	0.8359	0.9632	0.0001	0.0059	0.3424	0.0	0.0	0.6646	0.8269	0.5585	0.6586	0.2672	nan	0.3489	0.5512	0.0	0.7427	0.0	0.0	0.0	0.0	0.2798	0.0	nan	0.7027	0.0239	0.2868	0.2483	0.0	0.0	0.0	0.2342	0.0000	0.0	0.8203	0.6535	0.9096	0.0001	0.0041	0.2463	0.0	4

Framework versions

• Transformers 4.35.2
• TensorFlow 2.15.0
• Datasets 2.16.0
• Tokenizers 0.15.0