BEiT-b0-finetuned-metalography_V1_E30

This model is a fine-tuned version of ironchanchellor/BEiT-b0-finetuned-metalography_V1 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.0159
Mean Iou: 0.8022
Mean Accuracy: 0.9778
Overall Accuracy: 0.9954
Accuracy Background: nan
Accuracy Haz: 0.9959
Accuracy Matrix: 0.9882
Accuracy Porosity: 0.9275
Accuracy Carbides: 0.9786
Accuracy Substrate: 0.9987
Iou Background: 0.0
Iou Haz: 0.9940
Iou Matrix: 0.9774
Iou Porosity: 0.8864
Iou Carbides: 0.9595
Iou Substrate: 0.9960

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: 4
eval_batch_size: 4
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 20

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
3.0046	0.34	200	0.1249	0.7613	0.9391	0.9749	nan	0.9867	0.9781	0.8392	0.9193	0.9720	0.9435	0.9504	0.8045	0.9069	0.9626
0.049	0.68	400	0.0436	0.7770	0.9539	0.9874	nan	0.9888	0.9849	0.8361	0.9705	0.9892	0.9741	0.9704	0.7906	0.9458	0.9813
0.0669	1.02	600	0.0441	0.7755	0.9496	0.9871	nan	0.9776	0.9828	0.8194	0.9708	0.9974	0.9738	0.9682	0.7867	0.9427	0.9816
0.0413	1.36	800	0.0495	0.7610	0.9343	0.9858	nan	0.9913	0.9793	0.7380	0.9778	0.9849	0.9712	0.9678	0.7028	0.9451	0.9791
0.0433	1.69	1000	0.0313	0.7793	0.9565	0.9904	nan	0.9916	0.9772	0.8358	0.9841	0.9938	0.9831	0.9685	0.7920	0.9440	0.9881
0.0056	2.03	1200	0.0456	0.7774	0.9532	0.9872	nan	0.9748	0.9835	0.8362	0.9721	0.9992	0.9736	0.9697	0.7924	0.9470	0.9815
0.0291	2.37	1400	0.0863	0.7723	0.9461	0.9816	nan	0.9654	0.9830	0.8120	0.9759	0.9942	0.9574	0.9710	0.7871	0.9484	0.9697
0.0351	2.71	1600	0.0347	0.7722	0.9468	0.9883	nan	0.9839	0.9823	0.8123	0.9585	0.9968	0.9794	0.9631	0.7635	0.9417	0.9856
0.0469	3.05	1800	0.0279	0.7636	0.9289	0.9911	nan	0.9932	0.9869	0.7041	0.9663	0.9940	0.9848	0.9705	0.6897	0.9472	0.9893
0.031	3.39	2000	0.0371	0.7816	0.9602	0.9884	nan	0.9918	0.9815	0.8604	0.9784	0.9889	0.9770	0.9702	0.8120	0.9467	0.9835
0.0363	3.73	2200	0.0417	0.7662	0.9331	0.9885	nan	0.9826	0.9841	0.7276	0.9753	0.9960	0.9769	0.9715	0.7172	0.9479	0.9837
0.0123	4.07	2400	0.0467	0.7797	0.9594	0.9894	nan	0.9936	0.9830	0.8541	0.9766	0.9895	0.9794	0.9714	0.7923	0.9499	0.9852
0.0178	4.41	2600	0.0261	0.7827	0.9617	0.9917	nan	0.9944	0.9775	0.8584	0.9838	0.9943	0.9864	0.9694	0.8041	0.9459	0.9905
0.0179	4.75	2800	0.0254	0.7894	0.9654	0.9920	nan	0.9896	0.9856	0.8801	0.9740	0.9976	0.9862	0.9726	0.8360	0.9508	0.9904
0.0136	5.08	3000	0.0340	0.7851	0.9571	0.9900	nan	0.9839	0.9922	0.8573	0.9540	0.9983	0.9814	0.9697	0.8296	0.9428	0.9869
0.0144	5.42	3200	0.0383	0.7814	0.9533	0.9896	nan	0.9908	0.9893	0.8270	0.9675	0.9918	0.9797	0.9728	0.8001	0.9501	0.9855
0.0204	5.76	3400	0.0375	0.7852	0.9586	0.9899	nan	0.9819	0.9868	0.8528	0.9727	0.9987	0.9800	0.9729	0.8212	0.9511	0.9860
0.0178	6.1	3600	0.0217	0.7871	0.9645	0.9928	nan	0.9943	0.9839	0.8732	0.9753	0.9960	0.9888	0.9719	0.8201	0.9497	0.9923
0.0074	6.44	3800	0.0244	0.7883	0.9637	0.9921	nan	0.9923	0.9801	0.8672	0.9829	0.9960	0.9868	0.9713	0.8325	0.9487	0.9908
0.0285	6.78	4000	0.0271	0.7880	0.9642	0.9920	nan	0.9911	0.9830	0.8694	0.9812	0.9964	0.9861	0.9732	0.8272	0.9514	0.9903
0.0261	7.12	4200	0.0214	0.7861	0.9604	0.9933	nan	0.9937	0.9833	0.8477	0.9799	0.9973	0.9898	0.9730	0.8094	0.9515	0.9929
0.0261	7.46	4400	0.0219	0.7759	0.9411	0.9931	nan	0.9912	0.9862	0.7525	0.9771	0.9985	0.9891	0.9732	0.7479	0.9528	0.9925
0.0463	7.8	4600	0.0254	0.7930	0.9684	0.9927	nan	0.9911	0.9890	0.8961	0.9681	0.9977	0.9877	0.9734	0.8540	0.9514	0.9914
0.0108	8.14	4800	0.0225	0.7909	0.9688	0.9929	nan	0.9906	0.9893	0.9001	0.9655	0.9987	0.9888	0.9723	0.8413	0.9508	0.9923
4.1436	8.47	5000	0.0209	0.7910	0.9669	0.9935	nan	0.9919	0.9861	0.8817	0.9759	0.9986	0.9898	0.9741	0.8358	0.9531	0.9931
0.0311	8.81	5200	0.0248	0.7944	0.9734	0.9929	nan	0.9918	0.9838	0.9144	0.9795	0.9975	0.9880	0.9739	0.8591	0.9538	0.9918
0.0286	9.15	5400	0.0885	0.7883	0.9681	0.9848	nan	0.9680	0.9857	0.9103	0.9788	0.9979	0.9649	0.9751	0.8603	0.9545	0.9752
0.0065	9.49	5600	0.0211	0.7908	0.9681	0.9937	nan	0.9944	0.9806	0.8850	0.9830	0.9978	0.9911	0.9721	0.8359	0.9518	0.9940
0.0209	9.83	5800	0.0284	0.7951	0.9737	0.9927	nan	0.9943	0.9812	0.9139	0.9839	0.9955	0.9881	0.9726	0.8674	0.9510	0.9917
0.0109	10.17	6000	0.0210	0.7935	0.9678	0.9936	nan	0.9955	0.9862	0.8838	0.9777	0.9959	0.9898	0.9749	0.8489	0.9547	0.9930
0.0243	10.51	6200	0.0227	0.7943	0.9706	0.9934	nan	0.9958	0.9869	0.8993	0.9758	0.9954	0.9893	0.9748	0.8544	0.9550	0.9925
0.0138	10.85	6400	0.0195	0.7927	0.9701	0.9939	nan	0.9934	0.9846	0.8969	0.9771	0.9987	0.9911	0.9737	0.8431	0.9540	0.9939
0.0062	11.19	6600	0.0166	0.7978	0.9763	0.9948	nan	0.9969	0.9880	0.9237	0.9754	0.9972	0.9929	0.9754	0.8674	0.9557	0.9953
0.0325	11.53	6800	0.0175	0.7972	0.9770	0.9945	nan	0.9962	0.9864	0.9290	0.9759	0.9973	0.9922	0.9747	0.8659	0.9557	0.9949
0.0152	11.86	7000	0.0163	0.7969	0.9710	0.9948	nan	0.9966	0.9872	0.8976	0.9760	0.9977	0.9933	0.9750	0.8622	0.9555	0.9954
0.0165	12.2	7200	0.0179	0.7921	0.9656	0.9946	nan	0.9954	0.9843	0.8714	0.9783	0.9985	0.9929	0.9734	0.8364	0.9547	0.9954
0.0931	12.54	7400	0.0280	0.7929	0.9658	0.9926	nan	0.9896	0.9894	0.8851	0.9660	0.9989	0.9879	0.9727	0.8552	0.9502	0.9916
0.012	12.88	7600	0.0162	0.7975	0.9734	0.9951	nan	0.9964	0.9879	0.9093	0.9748	0.9983	0.9938	0.9758	0.8633	0.9561	0.9959
0.0223	13.22	7800	0.0786	0.7930	0.9691	0.9885	nan	0.9778	0.9877	0.9061	0.9759	0.9981	0.9751	0.9755	0.8688	0.9567	0.9823
0.0063	13.56	8000	0.0260	0.7969	0.9715	0.9933	nan	0.9907	0.9845	0.9024	0.9807	0.9990	0.9890	0.9745	0.8724	0.9529	0.9924
0.0245	13.9	8200	0.0361	0.7964	0.9744	0.9929	nan	0.9900	0.9878	0.9194	0.9765	0.9983	0.9876	0.9758	0.8669	0.9570	0.9913
0.0279	14.24	8400	0.0406	0.7958	0.9691	0.9925	nan	0.9882	0.9874	0.8934	0.9778	0.9987	0.9863	0.9757	0.8649	0.9575	0.9904
0.0142	14.58	8600	0.0241	0.7978	0.9727	0.9937	nan	0.9915	0.9877	0.9086	0.9768	0.9988	0.9896	0.9761	0.8712	0.9574	0.9928
0.0147	14.92	8800	0.0245	0.7967	0.9711	0.9937	nan	0.9918	0.9871	0.8997	0.9781	0.9987	0.9898	0.9759	0.8640	0.9572	0.9930
0.0031	15.25	9000	0.0154	0.8016	0.9784	0.9953	nan	0.9966	0.9870	0.9320	0.9781	0.9984	0.9942	0.9763	0.8857	0.9574	0.9962
0.0099	15.59	9200	0.0164	0.8007	0.9806	0.9951	nan	0.9955	0.9833	0.9403	0.9848	0.9990	0.9937	0.9756	0.8827	0.9565	0.9958
0.0076	15.93	9400	0.0168	0.7997	0.9750	0.9948	nan	0.9956	0.9876	0.9171	0.9767	0.9981	0.9927	0.9763	0.8762	0.9580	0.9951
0.0444	16.27	9600	0.0195	0.8007	0.9791	0.9946	nan	0.9945	0.9874	0.9371	0.9782	0.9984	0.9921	0.9764	0.8831	0.9580	0.9946
0.0147	16.61	9800	0.0179	0.8006	0.9792	0.9949	nan	0.9943	0.9879	0.9368	0.9778	0.9989	0.9926	0.9769	0.8805	0.9585	0.9950
0.0097	16.95	10000	0.0262	0.8005	0.9776	0.9938	nan	0.9908	0.9884	0.9331	0.9763	0.9993	0.9897	0.9767	0.8851	0.9584	0.9929
0.0261	17.29	10200	0.0187	0.8014	0.9805	0.9946	nan	0.9938	0.9878	0.9440	0.9779	0.9987	0.9920	0.9768	0.8866	0.9585	0.9945
0.0215	17.63	10400	0.0195	0.8010	0.9769	0.9947	nan	0.9945	0.9863	0.9239	0.9815	0.9984	0.9921	0.9769	0.8833	0.9586	0.9947
0.0187	17.97	10600	0.0156	0.8019	0.9809	0.9953	nan	0.9961	0.9865	0.9430	0.9804	0.9986	0.9940	0.9768	0.8858	0.9586	0.9960
0.0002	18.31	10800	0.0171	0.8018	0.9798	0.9951	nan	0.9952	0.9870	0.9382	0.9798	0.9987	0.9933	0.9769	0.8858	0.9590	0.9955
0.0079	18.64	11000	0.0169	0.8016	0.9787	0.9951	nan	0.9950	0.9875	0.9325	0.9794	0.9990	0.9934	0.9772	0.8846	0.9591	0.9956
0.0134	18.98	11200	0.0177	0.8022	0.9798	0.9949	nan	0.9946	0.9878	0.9383	0.9794	0.9987	0.9927	0.9773	0.8886	0.9593	0.9951
0.0054	19.32	11400	0.0166	0.8017	0.9783	0.9952	nan	0.9958	0.9874	0.9304	0.9794	0.9986	0.9936	0.9772	0.8845	0.9592	0.9957
0.0098	19.66	11600	0.0160	0.8021	0.9801	0.9954	nan	0.9960	0.9873	0.9386	0.9799	0.9987	0.9940	0.9772	0.8858	0.9592	0.9960
0.0236	20.0	11800	0.0159	0.8022	0.9778	0.9954	nan	0.9959	0.9882	0.9275	0.9786	0.9987	0.9940	0.9774	0.8864	0.9595	0.9960

Framework versions

Transformers 4.32.1
Pytorch 2.0.1+cu118
Datasets 2.14.4
Tokenizers 0.13.3

BEiT-b0-finetuned-metalography_V1_E30

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js