This model is a fine-tuned version of nvidia/mit-b2 on the dskong07/chargers-full-v0.1 dataset. It achieves the following results on the evaluation set:

Loss: 0.3562
Mean Iou: 0.7932
Mean Accuracy: 0.8741
Overall Accuracy: 0.9202
Accuracy Unlabeled: nan
Accuracy Screen: 0.8965
Accuracy Body: 0.9162
Accuracy Cable: 0.7137
Accuracy Plug: 0.9024
Accuracy Void-background: 0.9416
Iou Unlabeled: nan
Iou Screen: 0.7980
Iou Body: 0.7952
Iou Cable: 0.6210
Iou Plug: 0.8498
Iou Void-background: 0.9017

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: 16
eval_batch_size: 16
seed: 42
optimizer: Use OptimizerNames.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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Screen	Accuracy Body	Accuracy Cable	Accuracy Plug	Accuracy Void-background	Iou Unlabeled	Iou Screen	Iou Body	Iou Cable	Iou Plug	Iou Void-background
0.791	2.2222	20	0.9137	0.6056	0.7582	0.8259	nan	0.7897	0.9011	0.3986	0.8781	0.8237	nan	0.5894	0.6325	0.3127	0.7018	0.7917
0.5745	4.4444	40	0.5314	0.6800	0.7953	0.8728	nan	0.8035	0.8500	0.5621	0.8461	0.9150	nan	0.6561	0.7012	0.4305	0.7589	0.8532
0.299	6.6667	60	0.5239	0.7206	0.8270	0.8901	nan	0.8754	0.8941	0.5805	0.8698	0.9153	nan	0.7133	0.7331	0.5038	0.7821	0.8707
0.2347	8.8889	80	0.4256	0.7400	0.8361	0.8968	nan	0.8764	0.8594	0.6402	0.8687	0.9361	nan	0.7360	0.7410	0.5458	0.8023	0.8750
0.2687	11.1111	100	0.4235	0.7527	0.8554	0.9026	nan	0.9298	0.8578	0.6598	0.8930	0.9367	nan	0.7357	0.7511	0.5652	0.8284	0.8833
0.1953	13.3333	120	0.4096	0.7602	0.8623	0.9033	nan	0.8765	0.8807	0.7199	0.9064	0.9280	nan	0.7666	0.7590	0.5933	0.8012	0.8808
0.1998	15.5556	140	0.3897	0.7644	0.8591	0.9076	nan	0.8287	0.9400	0.6962	0.9132	0.9173	nan	0.7548	0.7751	0.5922	0.8125	0.8874
0.1636	17.7778	160	0.3818	0.7810	0.8675	0.9140	nan	0.8802	0.9041	0.7048	0.9117	0.9368	nan	0.7889	0.7802	0.6022	0.8398	0.8939
0.1554	20.0	180	0.3784	0.7718	0.8826	0.9119	nan	0.9614	0.8719	0.7156	0.9292	0.9348	nan	0.7299	0.7721	0.6117	0.8482	0.8972
0.1187	22.2222	200	0.3712	0.7851	0.8735	0.9159	nan	0.8778	0.9223	0.7164	0.9201	0.9310	nan	0.7982	0.7885	0.6086	0.8347	0.8957
0.1217	24.4444	220	0.3630	0.7898	0.8841	0.9174	nan	0.9005	0.9160	0.7352	0.9387	0.9301	nan	0.8053	0.7934	0.6209	0.8326	0.8969
0.092	26.6667	240	0.3711	0.7882	0.8801	0.9184	nan	0.9210	0.9023	0.7309	0.9067	0.9398	nan	0.7766	0.7899	0.6158	0.8574	0.9013
0.0906	28.8889	260	0.3732	0.7907	0.8798	0.9189	nan	0.8996	0.9158	0.7235	0.9247	0.9354	nan	0.7915	0.7934	0.6219	0.8466	0.9002
0.1034	31.1111	280	0.3899	0.7924	0.8801	0.9198	nan	0.9053	0.9124	0.7297	0.9141	0.9389	nan	0.7878	0.7927	0.6203	0.8586	0.9025
0.084	33.3333	300	0.3613	0.7925	0.8792	0.9200	nan	0.9141	0.9073	0.7117	0.9221	0.9406	nan	0.7925	0.7934	0.6163	0.8584	0.9020
0.1075	35.5556	320	0.3712	0.7901	0.8808	0.9193	nan	0.9249	0.9068	0.7126	0.9216	0.9384	nan	0.7866	0.7938	0.6175	0.8513	0.9013
0.1099	37.7778	340	0.4036	0.7934	0.8823	0.9203	nan	0.9055	0.9211	0.7249	0.9251	0.9350	nan	0.7972	0.7982	0.6235	0.8465	0.9018
0.2037	40.0	360	0.3782	0.7894	0.8719	0.9188	nan	0.8953	0.9155	0.6951	0.9145	0.9393	nan	0.7988	0.7936	0.6117	0.8433	0.8997
0.103	42.2222	380	0.3521	0.7921	0.8788	0.9199	nan	0.8977	0.9211	0.7155	0.9234	0.9361	nan	0.7992	0.7983	0.6195	0.8428	0.9010
0.0741	44.4444	400	0.3763	0.7922	0.8818	0.9199	nan	0.9081	0.9241	0.7135	0.9304	0.9330	nan	0.7982	0.7992	0.6216	0.8414	0.9008
0.0831	46.6667	420	0.3501	0.7921	0.8748	0.9199	nan	0.8859	0.9229	0.7177	0.9094	0.9382	nan	0.7934	0.7957	0.6227	0.8467	0.9018
0.079	48.8889	440	0.3562	0.7932	0.8741	0.9202	nan	0.8965	0.9162	0.7137	0.9024	0.9416	nan	0.7980	0.7952	0.6210	0.8498	0.9017

Framework versions

Transformers 4.48.3
Pytorch 2.6.0+cu124
Datasets 3.2.0
Tokenizers 0.21.0

irvingz/segformer-b2-finetuned-segments-chargers-full-v2.1

segformer-b1-finetuned-segments-chargers-full-v2.1

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

irvingz/
segformer-b2-finetuned-segments-chargers-full-v2.1