segformer-b0_DsB_B16

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

Loss: 0.0135
Mean Iou: 0.7956
Mean Accuracy: 0.9699
Overall Accuracy: 0.9936
Accuracy Background: nan
Accuracy Haz: 0.9976
Accuracy Matrix: 0.9895
Accuracy Porosity: 0.8969
Accuracy Carbides: 0.9682
Accuracy Substrate: 0.9974
Iou Background: 0.0
Iou Haz: 0.9951
Iou Matrix: 0.9758
Iou Porosity: 0.8600
Iou Carbides: 0.9476
Iou Substrate: 0.9952

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Haz	Accuracy Matrix	Accuracy Porosity	Accuracy Carbides	Accuracy Substrate	Iou Haz	Iou Matrix	Iou Porosity	Iou Carbides	Iou Substrate
0.3772	1.0	55	0.2719	0.5627	0.7173	0.9514	nan	0.9801	0.9565	0.0	0.6670	0.9831	0.9638	0.8389	0.0	0.6098	0.9638
0.2401	2.0	110	0.1641	0.5995	0.7615	0.9671	nan	0.9869	0.9149	0.0	0.9168	0.9890	0.9758	0.8802	0.0	0.7632	0.9776
0.1514	3.0	165	0.1051	0.6215	0.7679	0.9771	nan	0.9872	0.9785	0.0	0.8834	0.9907	0.9781	0.9285	0.0	0.8426	0.9795
0.2762	4.0	220	0.0833	0.6322	0.7786	0.9819	nan	0.9866	0.9788	0.0	0.9337	0.9939	0.9810	0.9464	0.0	0.8841	0.9819
0.0959	5.0	275	0.0679	0.6346	0.7799	0.9821	nan	0.9892	0.9801	0.0	0.9407	0.9895	0.9799	0.9512	0.0	0.8966	0.9799
0.0816	6.0	330	0.0583	0.6369	0.7809	0.9835	nan	0.9927	0.9800	0.0	0.9428	0.9892	0.9820	0.9546	0.0	0.9026	0.9820
0.0719	7.0	385	0.0494	0.6378	0.7809	0.9848	nan	0.9925	0.9878	0.0	0.9335	0.9908	0.9843	0.9560	0.0	0.9025	0.9842
0.0661	8.0	440	0.0432	0.6383	0.7861	0.9851	nan	0.9916	0.9705	0.0	0.9754	0.9928	0.9851	0.9571	0.0	0.9029	0.9847
0.0406	9.0	495	0.0381	0.6422	0.7843	0.9865	nan	0.9913	0.9873	0.0005	0.9481	0.9942	0.9861	0.9620	0.0005	0.9187	0.9856
0.0775	10.0	550	0.0351	0.6437	0.7854	0.9871	nan	0.9924	0.9890	0.0068	0.9438	0.9948	0.9873	0.9627	0.0068	0.9186	0.9868
0.0507	11.0	605	0.0322	0.6570	0.8019	0.9875	nan	0.9922	0.9802	0.0746	0.9673	0.9953	0.9871	0.9646	0.0746	0.9290	0.9869
0.0326	12.0	660	0.0301	0.6790	0.8265	0.9882	nan	0.9924	0.9893	0.2030	0.9530	0.9948	0.9878	0.9659	0.2030	0.9298	0.9875
0.0362	13.0	715	0.0399	0.7112	0.8693	0.9844	nan	0.9774	0.9816	0.4143	0.9741	0.9992	0.9764	0.9680	0.4143	0.9337	0.9748
0.0382	14.0	770	0.0266	0.7410	0.9005	0.9895	nan	0.9927	0.9858	0.5638	0.9648	0.9955	0.9886	0.9692	0.5634	0.9365	0.9882
0.0269	15.0	825	0.0259	0.7413	0.9001	0.9896	nan	0.9966	0.9892	0.5674	0.9556	0.9915	0.9885	0.9689	0.5673	0.9351	0.9880
0.0324	16.0	880	0.0246	0.7550	0.9172	0.9903	nan	0.9946	0.9859	0.6430	0.9678	0.9947	0.9896	0.9704	0.6428	0.9381	0.9892
0.0344	17.0	935	0.0232	0.7663	0.9312	0.9904	nan	0.9954	0.9842	0.7110	0.9720	0.9936	0.9897	0.9707	0.7086	0.9393	0.9893
0.0337	18.0	990	0.0233	0.7636	0.9260	0.9899	nan	0.9965	0.9918	0.7038	0.9454	0.9927	0.9900	0.9678	0.7026	0.9315	0.9896
0.0391	19.0	1045	0.0215	0.7713	0.9359	0.9912	nan	0.9961	0.9891	0.7365	0.9638	0.9943	0.9909	0.9720	0.7340	0.9403	0.9905
0.0324	20.0	1100	0.0214	0.7819	0.9507	0.9912	nan	0.9929	0.9883	0.8091	0.9660	0.9975	0.9903	0.9731	0.7963	0.9420	0.9899
0.0384	21.0	1155	0.0196	0.7835	0.9518	0.9916	nan	0.9947	0.9894	0.8168	0.9615	0.9968	0.9915	0.9726	0.8047	0.9412	0.9913
0.0289	22.0	1210	0.0191	0.7863	0.9569	0.9919	nan	0.9965	0.9851	0.8349	0.9725	0.9955	0.9922	0.9728	0.8182	0.9428	0.9920
0.0179	23.0	1265	0.0194	0.7870	0.9573	0.9916	nan	0.9978	0.9890	0.8426	0.9643	0.9929	0.9911	0.9734	0.8240	0.9426	0.9907
0.0269	24.0	1320	0.0183	0.7868	0.9572	0.9922	nan	0.9948	0.9873	0.8356	0.9713	0.9973	0.9921	0.9741	0.8186	0.9441	0.9919
0.0348	25.0	1375	0.0182	0.7898	0.9656	0.9919	nan	0.9951	0.9821	0.8749	0.9791	0.9970	0.9921	0.9727	0.8402	0.9421	0.9919
0.0389	26.0	1430	0.0169	0.7899	0.9625	0.9924	nan	0.9956	0.9879	0.8619	0.9702	0.9968	0.9927	0.9741	0.8352	0.9448	0.9927
0.0275	27.0	1485	0.0167	0.7910	0.9647	0.9925	nan	0.9961	0.9894	0.8762	0.9652	0.9967	0.9930	0.9742	0.8419	0.9441	0.9928
0.0217	28.0	1540	0.0165	0.7913	0.9672	0.9926	nan	0.9964	0.9867	0.8836	0.9726	0.9965	0.9930	0.9745	0.8421	0.9453	0.9929
0.0208	29.0	1595	0.0164	0.7920	0.9665	0.9925	nan	0.9972	0.9888	0.8839	0.9671	0.9954	0.9930	0.9743	0.8471	0.9450	0.9929
0.0255	30.0	1650	0.0159	0.7922	0.9655	0.9927	nan	0.9959	0.9897	0.8793	0.9652	0.9974	0.9934	0.9746	0.8470	0.9449	0.9934
0.0207	31.0	1705	0.0158	0.7908	0.9615	0.9929	nan	0.9963	0.9902	0.8584	0.9656	0.9971	0.9936	0.9748	0.8375	0.9451	0.9936
0.0221	32.0	1760	0.0157	0.7915	0.9650	0.9928	nan	0.9968	0.9843	0.8692	0.9782	0.9966	0.9936	0.9742	0.8426	0.9452	0.9935
0.0282	33.0	1815	0.0156	0.7927	0.9665	0.9927	nan	0.9957	0.9909	0.8868	0.9611	0.9979	0.9935	0.9743	0.8511	0.9439	0.9935
0.0169	34.0	1870	0.0149	0.7931	0.9665	0.9931	nan	0.9972	0.9890	0.8816	0.9679	0.9966	0.9940	0.9750	0.8493	0.9461	0.9941
0.0258	35.0	1925	0.0147	0.7939	0.9691	0.9932	nan	0.9964	0.9873	0.8919	0.9722	0.9978	0.9941	0.9752	0.8529	0.9471	0.9941
0.0211	36.0	1980	0.0148	0.7941	0.9686	0.9931	nan	0.9972	0.9883	0.8917	0.9693	0.9967	0.9940	0.9750	0.8546	0.9468	0.9941
0.0194	37.0	2035	0.0151	0.7938	0.9678	0.9929	nan	0.9982	0.9892	0.8883	0.9686	0.9949	0.9935	0.9753	0.8535	0.9469	0.9935
0.0135	38.0	2090	0.0145	0.7939	0.9692	0.9933	nan	0.9973	0.9853	0.8886	0.9776	0.9970	0.9944	0.9751	0.8527	0.9468	0.9944
0.1317	39.0	2145	0.0141	0.7938	0.9673	0.9932	nan	0.9974	0.9904	0.8878	0.9638	0.9970	0.9946	0.9749	0.8532	0.9457	0.9946
0.0204	40.0	2200	0.0140	0.7945	0.9689	0.9934	nan	0.9967	0.9877	0.8903	0.9722	0.9978	0.9945	0.9756	0.8545	0.9475	0.9946
0.0195	41.0	2255	0.0139	0.7948	0.9691	0.9934	nan	0.9975	0.9896	0.8941	0.9673	0.9971	0.9947	0.9756	0.8571	0.9469	0.9947
0.0141	42.0	2310	0.0143	0.7951	0.9695	0.9933	nan	0.9976	0.9896	0.8955	0.9683	0.9964	0.9943	0.9756	0.8585	0.9474	0.9945
0.0111	43.0	2365	0.0138	0.7949	0.9688	0.9935	nan	0.9973	0.9880	0.8886	0.9730	0.9973	0.9947	0.9758	0.8560	0.9480	0.9948
0.0205	44.0	2420	0.0138	0.7954	0.9695	0.9935	nan	0.9971	0.9881	0.8928	0.9721	0.9975	0.9948	0.9757	0.8589	0.9478	0.9949
0.0156	45.0	2475	0.0137	0.7951	0.9682	0.9936	nan	0.9974	0.9893	0.8874	0.9697	0.9974	0.9949	0.9759	0.8568	0.9478	0.9950
0.0187	46.0	2530	0.0135	0.7952	0.9686	0.9937	nan	0.9975	0.9888	0.8877	0.9715	0.9973	0.9950	0.9760	0.8569	0.9482	0.9951
0.0073	47.0	2585	0.0135	0.7957	0.9709	0.9936	nan	0.9975	0.9879	0.8996	0.9723	0.9974	0.9950	0.9758	0.8605	0.9481	0.9951
0.0136	48.0	2640	0.0133	0.7958	0.9711	0.9936	nan	0.9978	0.9877	0.9002	0.9726	0.9972	0.9951	0.9758	0.8605	0.9482	0.9951
0.0209	49.0	2695	0.0135	0.7957	0.9716	0.9935	nan	0.9975	0.9873	0.9034	0.9725	0.9975	0.9950	0.9755	0.8604	0.9480	0.9951
0.0161	50.0	2750	0.0135	0.7956	0.9699	0.9936	nan	0.9976	0.9895	0.8969	0.9682	0.9974	0.9951	0.9758	0.8600	0.9476	0.9952

Framework versions

Transformers 4.33.1
Pytorch 2.0.1+cu118
Datasets 2.14.5
Tokenizers 0.13.3

segformer-b0_DsB_B16

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js