BEiT-b0-finetuned-metalography_V1

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: 0.0534
Mean Iou: 0.7698
Mean Accuracy: 0.9392
Overall Accuracy: 0.9862
Accuracy Background: nan
Accuracy Haz: 0.9845
Accuracy Matrix: 0.9860
Accuracy Porosity: 0.7693
Accuracy Carbides: 0.9660
Accuracy Substrate: 0.9902
Iou Background: 0.0
Iou Haz: 0.9719
Iou Matrix: 0.9690
Iou Porosity: 0.7544
Iou Carbides: 0.9446
Iou Substrate: 0.9790

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: 8

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
0.3113	0.17	200	0.2631	0.6813	0.8560	0.9403	nan	0.9300	0.9661	0.5554	0.8787	0.9498	0.8633	0.9337	0.5396	0.8484	0.9027
0.1329	0.34	400	0.3384	0.6355	0.8242	0.9010	nan	0.9831	0.9765	0.4140	0.9300	0.8170	0.8041	0.9011	0.4062	0.8925	0.8094
0.1102	0.51	600	0.1319	0.6913	0.8736	0.9618	nan	0.9676	0.9445	0.5080	0.9887	0.9592	0.9227	0.9208	0.5033	0.8597	0.9412
0.0182	0.68	800	0.2033	0.7063	0.8890	0.9575	nan	0.9293	0.9404	0.6076	0.9873	0.9806	0.9055	0.9344	0.5798	0.8880	0.9304
1.1404	0.85	1000	0.1376	0.7193	0.8875	0.9693	nan	0.9771	0.9856	0.5929	0.9157	0.9663	0.9348	0.9473	0.5812	0.9024	0.9503
1.673	1.02	1200	0.2288	0.7355	0.9290	0.9550	nan	0.9842	0.9694	0.7899	0.9746	0.9271	0.8997	0.9528	0.7220	0.9214	0.9168
0.1769	1.19	1400	0.3643	0.6840	0.8676	0.9132	nan	0.7880	0.9777	0.6237	0.9594	0.9894	0.7768	0.9581	0.5896	0.9289	0.8507
0.1443	1.35	1600	0.1294	0.7224	0.8898	0.9733	nan	0.9780	0.9908	0.5803	0.9283	0.9715	0.9439	0.9517	0.5662	0.9167	0.9561
0.0853	1.52	1800	0.1969	0.7185	0.8860	0.9633	nan	0.9255	0.9910	0.6207	0.8986	0.9940	0.9179	0.9438	0.6166	0.8922	0.9406
0.016	1.69	2000	0.3721	0.7342	0.9092	0.9627	nan	0.9242	0.9666	0.6944	0.9695	0.9913	0.9133	0.9517	0.6838	0.9205	0.9358
0.0799	1.86	2200	0.0765	0.7476	0.9155	0.9811	nan	0.9781	0.9821	0.6674	0.9639	0.9858	0.9603	0.9622	0.6607	0.9322	0.9704
0.0484	2.03	2400	0.1077	0.7449	0.9166	0.9738	nan	0.9741	0.9799	0.6936	0.9614	0.9742	0.9419	0.9639	0.6774	0.9307	0.9554
5.8301	2.2	2600	0.1588	0.7200	0.8966	0.9671	nan	0.9642	0.9474	0.6158	0.9824	0.9729	0.9315	0.9399	0.5986	0.9022	0.9475
0.0051	2.37	2800	0.1625	0.7429	0.9194	0.9670	nan	0.9399	0.9698	0.7236	0.9770	0.9866	0.9234	0.9592	0.7050	0.9262	0.9435
0.0722	2.54	3000	0.1039	0.7401	0.9078	0.9776	nan	0.9917	0.9771	0.6332	0.9683	0.9686	0.9523	0.9615	0.6296	0.9340	0.9631
0.0318	2.71	3200	0.1010	0.7583	0.9326	0.9759	nan	0.9688	0.9798	0.7620	0.9713	0.9813	0.9455	0.9643	0.7433	0.9373	0.9591
0.0154	2.88	3400	0.1191	0.7517	0.9274	0.9760	nan	0.9701	0.9693	0.7312	0.9848	0.9817	0.9472	0.9606	0.7123	0.9303	0.9600
0.0788	3.05	3600	0.0925	0.7553	0.9391	0.9790	nan	0.9736	0.9493	0.7925	0.9908	0.9893	0.9593	0.9448	0.7528	0.9053	0.9699
0.1632	3.22	3800	0.0994	0.7485	0.9171	0.9775	nan	0.9869	0.9819	0.6819	0.9633	0.9714	0.9505	0.9641	0.6774	0.9369	0.9620
0.8963	3.39	4000	0.1258	0.7515	0.9287	0.9673	nan	0.9711	0.9781	0.7623	0.9701	0.9619	0.9251	0.9634	0.7444	0.9348	0.9414
0.0082	3.56	4200	0.1316	0.7540	0.9392	0.9745	nan	0.9841	0.9621	0.7948	0.9860	0.9690	0.9456	0.9554	0.7435	0.9216	0.9576
0.2648	3.73	4400	0.0968	0.7631	0.9364	0.9799	nan	0.9708	0.9812	0.7741	0.9678	0.9884	0.9559	0.9665	0.7498	0.9393	0.9670
0.0396	3.9	4600	0.0703	0.7658	0.9362	0.9836	nan	0.9737	0.9848	0.7721	0.9561	0.9946	0.9662	0.9649	0.7538	0.9349	0.9751
0.0571	4.06	4800	0.0970	0.7613	0.9357	0.9798	nan	0.9869	0.9827	0.7649	0.9686	0.9753	0.9562	0.9663	0.7391	0.9404	0.9660
0.0316	4.23	5000	0.1017	0.7450	0.9128	0.9769	nan	0.9582	0.9756	0.6625	0.9749	0.9926	0.9483	0.9631	0.6561	0.9403	0.9619
0.0789	4.4	5200	0.0719	0.7679	0.9416	0.9831	nan	0.9765	0.9775	0.7851	0.9781	0.9906	0.9643	0.9665	0.7637	0.9394	0.9735
0.0689	4.57	5400	0.0739	0.7601	0.9289	0.9823	nan	0.9864	0.9851	0.7320	0.9592	0.9817	0.9629	0.9659	0.7206	0.9396	0.9719
0.0886	4.74	5600	0.1535	0.7577	0.9277	0.9778	nan	0.9587	0.9858	0.7367	0.9647	0.9926	0.9495	0.9669	0.7257	0.9414	0.9627
0.0887	4.91	5800	0.0821	0.7646	0.9358	0.9814	nan	0.9916	0.9876	0.7630	0.9619	0.9748	0.9598	0.9682	0.7479	0.9424	0.9690
0.0015	5.08	6000	0.0604	0.7617	0.9297	0.9843	nan	0.9857	0.9832	0.7222	0.9720	0.9854	0.9667	0.9683	0.7154	0.9449	0.9751
0.1276	5.25	6200	0.0907	0.7591	0.9300	0.9791	nan	0.9732	0.9824	0.7404	0.9697	0.9844	0.9535	0.9672	0.7270	0.9418	0.9651
0.0356	5.42	6400	0.1027	0.7541	0.9275	0.9749	nan	0.9938	0.9810	0.7298	0.9736	0.9594	0.9438	0.9671	0.7152	0.9432	0.9554
0.0359	5.59	6600	0.1316	0.7561	0.9244	0.9790	nan	0.9625	0.9860	0.7161	0.9654	0.9921	0.9524	0.9682	0.7078	0.9436	0.9648
0.0085	5.76	6800	0.0660	0.7715	0.9442	0.9859	nan	0.9834	0.9788	0.7914	0.9762	0.9910	0.9717	0.9679	0.7676	0.9427	0.9789
0.0351	5.93	7000	0.0609	0.7633	0.9317	0.9843	nan	0.9797	0.9758	0.7367	0.9748	0.9916	0.9689	0.9639	0.7286	0.9417	0.9768
0.0185	6.1	7200	0.0520	0.7718	0.9442	0.9857	nan	0.9923	0.9833	0.7905	0.9720	0.9831	0.9707	0.9690	0.7687	0.9445	0.9777
0.51	6.27	7400	0.0642	0.7717	0.9435	0.9846	nan	0.9746	0.9826	0.7921	0.9741	0.9943	0.9671	0.9692	0.7729	0.9451	0.9757
0.0246	6.44	7600	0.0608	0.7683	0.9389	0.9849	nan	0.9879	0.9831	0.7658	0.9727	0.9848	0.9686	0.9688	0.7516	0.9445	0.9762
0.1015	6.6	7800	0.0408	0.7716	0.9408	0.9882	nan	0.9901	0.9859	0.7704	0.9679	0.9900	0.9771	0.9699	0.7545	0.9452	0.9829
0.0598	6.77	8000	0.0454	0.7639	0.9319	0.9876	nan	0.9871	0.9823	0.7271	0.9713	0.9917	0.9764	0.9688	0.7117	0.9441	0.9824
0.002	6.94	8200	0.0543	0.7702	0.9398	0.9864	nan	0.9819	0.9883	0.7810	0.9544	0.9936	0.9735	0.9657	0.7636	0.9382	0.9804
0.0021	7.11	8400	0.0827	0.7736	0.9449	0.9873	nan	0.9819	0.9831	0.7916	0.9735	0.9944	0.9745	0.9697	0.7698	0.9463	0.9812
0.0008	7.28	8600	0.0530	0.7691	0.9391	0.9865	nan	0.9843	0.9808	0.7629	0.9764	0.9912	0.9728	0.9687	0.7487	0.9447	0.9798
0.0113	7.45	8800	0.0486	0.7696	0.9382	0.9876	nan	0.9889	0.9849	0.7578	0.9694	0.9898	0.9756	0.9689	0.7464	0.9451	0.9819
0.0014	7.62	9000	0.0553	0.7743	0.9476	0.9863	nan	0.9841	0.9825	0.8066	0.9740	0.9908	0.9720	0.9696	0.7788	0.9464	0.9792
0.0474	7.79	9200	0.0442	0.7704	0.9393	0.9872	nan	0.9890	0.9862	0.7668	0.9654	0.9891	0.9747	0.9686	0.7541	0.9439	0.9811
0.0278	7.96	9400	0.0534	0.7698	0.9392	0.9862	nan	0.9845	0.9860	0.7693	0.9660	0.9902	0.9719	0.9690	0.7544	0.9446	0.9790

Framework versions

Transformers 4.32.0
Pytorch 2.0.1+cu118
Datasets 2.14.4
Tokenizers 0.13.3

BEiT-b0-finetuned-metalography_V1

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js