segformer-b0_DsB_B32

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: 0.0214
Mean Iou: 0.7857
Mean Accuracy: 0.9582
Overall Accuracy: 0.9919
Accuracy Background: nan
Accuracy Haz: 0.9961
Accuracy Matrix: 0.9873
Accuracy Porosity: 0.8442
Accuracy Carbides: 0.9673
Accuracy Substrate: 0.9958
Iou Background: 0.0
Iou Haz: 0.9922
Iou Matrix: 0.9731
Iou Porosity: 0.8151
Iou Carbides: 0.9418
Iou Substrate: 0.9920

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: 32
eval_batch_size: 32
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
0.5872	1.0	28	0.6058	0.4664	0.6217	0.9104	nan	0.9664	0.9848	0.0	0.1885	0.9688	0.9370	0.7385	0.0	0.1872	0.9356
0.4248	2.0	56	0.2621	0.5819	0.7389	0.9589	nan	0.9868	0.9408	0.0	0.7871	0.9796	0.9665	0.8629	0.0	0.6949	0.9670
0.2915	3.0	84	0.1987	0.6000	0.7556	0.9682	nan	0.9825	0.9479	0.0	0.8547	0.9929	0.9753	0.8878	0.0	0.7589	0.9779
0.4211	4.0	112	0.1711	0.6091	0.7608	0.9718	nan	0.9842	0.9558	0.0	0.8703	0.9938	0.9774	0.9028	0.0	0.7947	0.9798
0.1881	5.0	140	0.1364	0.6228	0.7746	0.9781	nan	0.9867	0.9599	0.0	0.9332	0.9931	0.9799	0.9277	0.0	0.8473	0.9821
0.1595	6.0	168	0.1128	0.6278	0.7736	0.9798	nan	0.9893	0.9790	0.0	0.9097	0.9899	0.9788	0.9397	0.0	0.8682	0.9800
0.1218	7.0	196	0.0980	0.6317	0.7766	0.9821	nan	0.9884	0.9819	0.0	0.9190	0.9939	0.9818	0.9456	0.0	0.8796	0.9829
0.1338	8.0	224	0.0843	0.6338	0.7784	0.9833	nan	0.9905	0.9825	0.0	0.9257	0.9933	0.9834	0.9491	0.0	0.8858	0.9842
0.092	9.0	252	0.0771	0.6343	0.7816	0.9827	nan	0.9959	0.9711	0.0	0.9558	0.9851	0.9814	0.9507	0.0	0.8922	0.9814
0.1415	10.0	280	0.0662	0.6364	0.7816	0.9839	nan	0.9931	0.9774	0.0	0.9471	0.9902	0.9828	0.9538	0.0	0.8984	0.9832
0.0905	11.0	308	0.0613	0.6373	0.7819	0.9839	nan	0.9956	0.9770	0.0	0.9501	0.9870	0.9827	0.9550	0.0	0.9034	0.9828
0.078	12.0	336	0.0560	0.6387	0.7819	0.9855	nan	0.9925	0.9860	0.0	0.9383	0.9927	0.9858	0.9579	0.0	0.9028	0.9856
0.0557	13.0	364	0.0543	0.6391	0.7841	0.9848	nan	0.9856	0.9795	0.0	0.9576	0.9979	0.9829	0.9595	0.0	0.9100	0.9825
0.0819	14.0	392	0.0503	0.6397	0.7857	0.9855	nan	0.9935	0.9731	0.0	0.9704	0.9917	0.9857	0.9588	0.0	0.9087	0.9853
0.055	15.0	420	0.0462	0.6410	0.7862	0.9863	nan	0.9945	0.9777	0.0	0.9682	0.9907	0.9859	0.9618	0.0	0.9128	0.9853
0.0569	16.0	448	0.0430	0.6415	0.7841	0.9864	nan	0.9901	0.9865	0.0004	0.9479	0.9958	0.9860	0.9628	0.0004	0.9142	0.9856
0.0636	17.0	476	0.0418	0.6436	0.7879	0.9870	nan	0.9947	0.9802	0.0061	0.9670	0.9913	0.9870	0.9635	0.0061	0.9181	0.9866
0.0584	18.0	504	0.0397	0.6450	0.7876	0.9871	nan	0.9945	0.9869	0.0129	0.9523	0.9914	0.9868	0.9646	0.0129	0.9195	0.9862
0.1067	19.0	532	0.0407	0.6515	0.7967	0.9868	nan	0.9971	0.9825	0.0507	0.9665	0.9866	0.9850	0.9661	0.0507	0.9232	0.9840
0.0791	20.0	560	0.0369	0.6594	0.8050	0.9875	nan	0.9893	0.9866	0.0933	0.9585	0.9971	0.9863	0.9672	0.0933	0.9241	0.9857
0.081	21.0	588	0.0359	0.6918	0.8457	0.9874	nan	0.9970	0.9772	0.2902	0.9760	0.9883	0.9862	0.9654	0.2900	0.9236	0.9855
0.0507	22.0	616	0.0338	0.7100	0.8643	0.9887	nan	0.9961	0.9871	0.3886	0.9589	0.9909	0.9878	0.9678	0.3885	0.9288	0.9872
0.0336	23.0	644	0.0333	0.7162	0.8725	0.9886	nan	0.9897	0.9856	0.4257	0.9641	0.9973	0.9868	0.9690	0.4253	0.9296	0.9862
0.0515	24.0	672	0.0318	0.7374	0.8960	0.9895	nan	0.9933	0.9903	0.5503	0.9505	0.9955	0.9889	0.9679	0.5499	0.9293	0.9885
0.0722	25.0	700	0.0303	0.7384	0.8979	0.9898	nan	0.9957	0.9853	0.5497	0.9659	0.9930	0.9891	0.9695	0.5488	0.9340	0.9888
0.0853	26.0	728	0.0296	0.7499	0.9124	0.9897	nan	0.9913	0.9840	0.6189	0.9705	0.9972	0.9886	0.9698	0.6175	0.9355	0.9883
0.0603	27.0	756	0.0284	0.7531	0.9154	0.9901	nan	0.9950	0.9849	0.6351	0.9678	0.9941	0.9897	0.9698	0.6329	0.9365	0.9895
0.0319	28.0	784	0.0278	0.7648	0.9296	0.9905	nan	0.9948	0.9854	0.7053	0.9675	0.9948	0.9899	0.9705	0.7022	0.9369	0.9895
0.0314	29.0	812	0.0272	0.7701	0.9365	0.9905	nan	0.9946	0.9838	0.7379	0.9708	0.9952	0.9900	0.9704	0.7329	0.9372	0.9897
0.0453	30.0	840	0.0274	0.7706	0.9365	0.9902	nan	0.9971	0.9860	0.7424	0.9657	0.9913	0.9891	0.9707	0.7375	0.9379	0.9886
0.036	31.0	868	0.0260	0.7709	0.9362	0.9908	nan	0.9937	0.9885	0.7401	0.9623	0.9964	0.9903	0.9713	0.7360	0.9379	0.9900
0.0453	32.0	896	0.0254	0.7735	0.9401	0.9909	nan	0.9934	0.9864	0.7559	0.9681	0.9967	0.9904	0.9716	0.7501	0.9389	0.9901
0.0395	33.0	924	0.0252	0.7787	0.9483	0.9907	nan	0.9933	0.9835	0.7958	0.9722	0.9967	0.9901	0.9710	0.7827	0.9388	0.9897
0.0328	34.0	952	0.0242	0.7787	0.9470	0.9913	nan	0.9952	0.9869	0.7902	0.9670	0.9958	0.9913	0.9720	0.7784	0.9396	0.9910
0.0485	35.0	980	0.0237	0.7791	0.9473	0.9913	nan	0.9946	0.9872	0.7925	0.9655	0.9966	0.9911	0.9720	0.7805	0.9398	0.9909
0.0346	36.0	1008	0.0236	0.7827	0.9533	0.9913	nan	0.9949	0.9869	0.8230	0.9652	0.9963	0.9913	0.9718	0.8019	0.9403	0.9912
0.0328	37.0	1036	0.0232	0.7822	0.9517	0.9916	nan	0.9959	0.9884	0.8151	0.9640	0.9953	0.9916	0.9724	0.7976	0.9403	0.9914
0.0354	38.0	1064	0.0236	0.7824	0.9531	0.9912	nan	0.9974	0.9869	0.8212	0.9676	0.9927	0.9907	0.9725	0.8001	0.9409	0.9903
0.2898	39.0	1092	0.0222	0.7829	0.9528	0.9915	nan	0.9964	0.9885	0.8216	0.9627	0.9947	0.9915	0.9723	0.8021	0.9401	0.9913
0.0363	40.0	1120	0.0224	0.7835	0.9546	0.9915	nan	0.9960	0.9866	0.8272	0.9679	0.9953	0.9915	0.9725	0.8046	0.9411	0.9913
0.0387	41.0	1148	0.0222	0.7836	0.9555	0.9916	nan	0.9967	0.9849	0.8291	0.9724	0.9946	0.9916	0.9725	0.8052	0.9411	0.9914
0.0281	42.0	1176	0.0231	0.7845	0.9568	0.9913	nan	0.9972	0.9868	0.8395	0.9676	0.9930	0.9909	0.9725	0.8120	0.9411	0.9906
0.0172	43.0	1204	0.0223	0.7841	0.9559	0.9917	nan	0.9967	0.9868	0.8323	0.9691	0.9945	0.9917	0.9730	0.8072	0.9416	0.9914
0.0386	44.0	1232	0.0222	0.7853	0.9579	0.9917	nan	0.9965	0.9866	0.8429	0.9688	0.9948	0.9918	0.9728	0.8142	0.9416	0.9916
0.0332	45.0	1260	0.0219	0.7844	0.9559	0.9919	nan	0.9963	0.9872	0.8322	0.9685	0.9955	0.9920	0.9731	0.8079	0.9418	0.9918
0.0293	46.0	1288	0.0216	0.7845	0.9563	0.9920	nan	0.9959	0.9868	0.8331	0.9695	0.9960	0.9921	0.9732	0.8076	0.9420	0.9919
0.02	47.0	1316	0.0215	0.7855	0.9584	0.9919	nan	0.9961	0.9862	0.8438	0.9703	0.9957	0.9921	0.9730	0.8140	0.9419	0.9919
0.0245	48.0	1344	0.0214	0.7857	0.9589	0.9919	nan	0.9962	0.9857	0.8463	0.9707	0.9958	0.9921	0.9729	0.8155	0.9419	0.9920
0.0547	49.0	1372	0.0215	0.7858	0.9590	0.9917	nan	0.9963	0.9849	0.8473	0.9712	0.9955	0.9921	0.9724	0.8170	0.9416	0.9919
0.0273	50.0	1400	0.0214	0.7857	0.9582	0.9919	nan	0.9961	0.9873	0.8442	0.9673	0.9958	0.9922	0.9731	0.8151	0.9418	0.9920

Framework versions

Transformers 4.33.1
Pytorch 2.0.1+cu118
Datasets 2.14.5
Tokenizers 0.13.3

segformer-b0_DsB_B32

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js