segformer-b0-finetuned-segments-sidewalk-oct-22

This model is a fine-tuned version of nvidia/mit-b0 on the Saad287/SIXRay_Gun dataset. It achieves the following results on the evaluation set:

• Loss: 0.0727
• Mean Iou: 0.1716
• Mean Accuracy: 0.2272
• Overall Accuracy: 0.5822
• Accuracy No-label: nan
• Accuracy Object1: 0.6917
• Accuracy Object2: 0.5239
• Accuracy Object3: 0.0778
• Accuracy Object4: 0.0696
• Accuracy Object5: 0.0
• Accuracy Object6: 0.0
• Iou No-label: 0.0
• Iou Object1: 0.5988
• Iou Object2: 0.4586
• Iou Object3: 0.0758
• Iou Object4: 0.0684
• Iou Object5: 0.0
• Iou Object6: 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: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy No-label	Accuracy Object1	Accuracy Object2	Accuracy Object3	Accuracy Object4	Accuracy Object5	Accuracy Object6	Iou No-label	Iou Object1	Iou Object2	Iou Object3	Iou Object4	Iou Object5	Iou Object6
0.1266	0.0521	20	0.1356	0.0943	0.1241	0.3893	nan	0.5262	0.2186	0.0	0.0	0.0	0.0	0.0	0.4513	0.2089	0.0	0.0	0.0	0.0
0.1092	0.1042	40	0.1287	0.1213	0.1733	0.4985	nan	0.6110	0.4286	0.0	0.0	0.0	0.0	0.0	0.5087	0.3405	0.0	0.0	0.0	0.0
0.2208	0.1562	60	0.1160	0.1007	0.1384	0.4361	nan	0.5925	0.2378	0.0	0.0	0.0	0.0	0.0	0.4796	0.2253	0.0	0.0	0.0	0.0
0.0722	0.2083	80	0.1101	0.1112	0.1556	0.4948	nan	0.6781	0.2555	0.0	0.0	0.0	0.0	0.0	0.5337	0.2444	0.0	0.0	0.0	0.0
0.1476	0.2604	100	0.1162	0.0784	0.1115	0.3905	nan	0.5849	0.0840	0.0	0.0	0.0	0.0	0.0	0.4653	0.0832	0.0	0.0	0.0	0.0
0.1069	0.3125	120	0.1134	0.0970	0.1300	0.3504	nan	0.3937	0.3863	0.0	0.0	0.0	0.0	0.0	0.3681	0.3108	0.0	0.0	0.0	0.0
0.1365	0.3646	140	0.1155	0.1276	0.1826	0.5713	nan	0.7701	0.3256	0.0	0.0	0.0	0.0	0.0	0.5870	0.3061	0.0	0.0	0.0	0.0
0.0918	0.4167	160	0.1053	0.1279	0.1801	0.5374	nan	0.6877	0.3932	0.0	0.0	0.0	0.0	0.0	0.5440	0.3513	0.0	0.0	0.0	0.0
0.109	0.4688	180	0.1014	0.1200	0.1647	0.5025	nan	0.6596	0.3285	0.0	0.0	0.0	0.0	0.0	0.5336	0.3065	0.0	0.0	0.0	0.0
0.0906	0.5208	200	0.0959	0.1338	0.1851	0.5505	nan	0.7024	0.4078	0.0002	0.0	0.0	0.0	0.0	0.5734	0.3628	0.0002	0.0	0.0	0.0
0.0735	0.5729	220	0.0988	0.1066	0.1552	0.5157	nan	0.7370	0.1944	0.0	0.0	0.0	0.0	0.0	0.5557	0.1906	0.0	0.0	0.0	0.0
0.0534	0.625	240	0.0944	0.1308	0.1779	0.5155	nan	0.6374	0.4301	0.0	0.0	0.0	0.0	0.0	0.5435	0.3721	0.0	0.0	0.0	0.0
0.0646	0.6771	260	0.0888	0.1316	0.1782	0.5113	nan	0.6243	0.4451	0.0	0.0	0.0	0.0	0.0	0.5378	0.3836	0.0	0.0	0.0	0.0
0.0918	0.7292	280	0.0915	0.1361	0.2036	0.5495	nan	0.6184	0.6031	0.0002	0.0	0.0	0.0	0.0	0.5300	0.4227	0.0002	0.0	0.0	0.0
0.0408	0.7812	300	0.0930	0.1228	0.1654	0.4673	nan	0.5597	0.4330	0.0	0.0	0.0	0.0	0.0	0.4926	0.3672	0.0	0.0	0.0	0.0
0.0592	0.8333	320	0.0902	0.1175	0.1864	0.4581	nan	0.4423	0.6763	0.0	0.0	0.0	0.0	0.0	0.4111	0.4118	0.0	0.0	0.0	0.0
0.1232	0.8854	340	0.0865	0.1332	0.1826	0.5524	nan	0.7218	0.3685	0.0006	0.0047	0.0	0.0	0.0	0.5813	0.3459	0.0006	0.0047	0.0	0.0
0.0886	0.9375	360	0.0832	0.1409	0.1895	0.5316	nan	0.6377	0.4882	0.0029	0.0081	0.0	0.0	0.0	0.5571	0.4187	0.0029	0.0080	0.0	0.0
0.1017	0.9896	380	0.0834	0.1538	0.2141	0.6076	nan	0.7507	0.5053	0.0002	0.0286	0.0	0.0	0.0	0.6136	0.4339	0.0002	0.0285	0.0	0.0
0.0752	1.0417	400	0.0796	0.1454	0.1951	0.5326	nan	0.6349	0.4953	0.0056	0.0348	0.0	0.0	0.0	0.5505	0.4279	0.0056	0.0339	0.0	0.0
0.1932	1.0938	420	0.0791	0.1678	0.2318	0.5991	nan	0.6897	0.6074	0.0223	0.0716	0.0	0.0	0.0	0.5951	0.4879	0.0222	0.0698	0.0	0.0
0.0981	1.1458	440	0.0821	0.1434	0.1909	0.5522	nan	0.7081	0.3933	0.0162	0.0276	0.0	0.0	0.0	0.5899	0.3721	0.0160	0.0258	0.0	0.0
0.0492	1.1979	460	0.0778	0.1562	0.2103	0.5857	nan	0.7383	0.4459	0.0083	0.0693	0.0	0.0	0.0	0.6071	0.4113	0.0083	0.0666	0.0	0.0
0.0509	1.25	480	0.0793	0.1534	0.2028	0.5267	nan	0.6092	0.5256	0.0293	0.0530	0.0	0.0	0.0	0.5463	0.4500	0.0289	0.0483	0.0	0.0
0.0687	1.3021	500	0.0784	0.1848	0.2528	0.6293	nan	0.7450	0.5812	0.0154	0.1753	0.0	0.0	0.0	0.6296	0.4799	0.0154	0.1688	0.0	0.0
0.0731	1.3542	520	0.0785	0.1379	0.1795	0.5113	nan	0.6292	0.4299	0.0118	0.0060	0.0	0.0	0.0	0.5519	0.3956	0.0118	0.0060	0.0	0.0
0.037	1.4062	540	0.0767	0.1712	0.2348	0.6052	nan	0.6978	0.6093	0.0298	0.0718	0.0	0.0	0.0	0.6092	0.4886	0.0297	0.0709	0.0	0.0
0.0656	1.4583	560	0.0765	0.1693	0.2280	0.5944	nan	0.6972	0.5661	0.0505	0.0546	0.0	0.0	0.0	0.6022	0.4787	0.0502	0.0543	0.0	0.0
0.1244	1.5104	580	0.0750	0.1580	0.2096	0.5554	nan	0.6455	0.5464	0.0458	0.0200	0.0	0.0	0.0	0.5717	0.4693	0.0451	0.0198	0.0	0.0
0.0528	1.5625	600	0.0748	0.1827	0.2448	0.6157	nan	0.7343	0.5472	0.0655	0.1219	0.0	0.0	0.0	0.6238	0.4736	0.0642	0.1169	0.0	0.0
0.0818	1.6146	620	0.0733	0.1749	0.2370	0.5908	nan	0.6649	0.6280	0.0422	0.0868	0.0	0.0	0.0	0.5926	0.5054	0.0419	0.0846	0.0	0.0
0.0272	1.6667	640	0.0728	0.1772	0.2375	0.6001	nan	0.7048	0.5615	0.0640	0.0946	0.0	0.0	0.0	0.6094	0.4767	0.0630	0.0913	0.0	0.0
0.0358	1.7188	660	0.0734	0.1704	0.2254	0.5721	nan	0.6660	0.5484	0.0747	0.0634	0.0	0.0	0.0	0.5852	0.4709	0.0736	0.0628	0.0	0.0
0.0808	1.7708	680	0.0723	0.1761	0.2321	0.5862	nan	0.7035	0.5020	0.1054	0.0815	0.0	0.0	0.0	0.6020	0.4478	0.1022	0.0805	0.0	0.0
0.0663	1.8229	700	0.0721	0.1784	0.2361	0.5952	nan	0.7086	0.5257	0.0974	0.0852	0.0	0.0	0.0	0.6091	0.4612	0.0946	0.0839	0.0	0.0
0.0521	1.875	720	0.0720	0.1798	0.2392	0.5963	nan	0.6937	0.5680	0.0861	0.0877	0.0	0.0	0.0	0.6054	0.4824	0.0845	0.0864	0.0	0.0
0.0557	1.9271	740	0.0725	0.1758	0.2341	0.5959	nan	0.7035	0.5469	0.0758	0.0784	0.0	0.0	0.0	0.6081	0.4711	0.0744	0.0772	0.0	0.0
0.048	1.9792	760	0.0727	0.1716	0.2272	0.5822	nan	0.6917	0.5239	0.0778	0.0696	0.0	0.0	0.0	0.5988	0.4586	0.0758	0.0684	0.0	0.0

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1