SWIN_Gaudi_100

This model is a fine-tuned version of microsoft/swinv2-large-patch4-window12-192-22k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 6.5510
• Accuracy: 0.1798
• Memory Allocated (gb): 2.43
• Max Memory Allocated (gb): 21.35
• Total Memory Available (gb): 94.62

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: 0.001
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• total_train_batch_size: 512
• total_eval_batch_size: 512
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-06
• lr_scheduler_type: linear
• num_epochs: 100.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Allocated (gb)	Memory Allocated (gb)	Memory Available (gb)
4.0897	1.0	1313	6.5606	0.0689	2.46	21.24	94.62
2.8297	2.0	2626	6.3807	0.0806	2.46	21.25	94.62
2.3951	3.0	3939	6.4208	0.0884	2.46	21.25	94.62
2.1003	4.0	5252	6.1807	0.1014	2.46	21.25	94.62
1.93	5.0	6565	6.3296	0.1027	2.46	21.25	94.62
1.7486	6.0	7878	6.2377	0.1091	2.46	21.25	94.62
1.6498	7.0	9191	6.4809	0.1056	2.46	21.25	94.62
1.5722	8.0	10504	6.2681	0.1150	2.46	21.25	94.62
1.458	9.0	11817	6.2907	0.1137	2.46	21.25	94.62
1.4131	10.0	13130	6.4250	0.1149	2.46	21.25	94.62
1.3132	11.0	14443	6.4233	0.1148	2.46	21.28	94.62
1.2835	12.0	15756	6.3784	0.1192	2.46	21.28	94.62
1.2414	13.0	17069	6.4416	0.1161	2.46	21.28	94.62
1.1652	14.0	18382	6.5069	0.1194	2.46	21.28	94.62
1.1415	15.0	19695	6.3847	0.1265	2.46	21.28	94.62
1.118	16.0	21008	6.3110	0.1265	2.46	21.28	94.62
1.065	17.0	22321	6.4024	0.1283	2.46	21.28	94.62
1.0469	18.0	23634	6.1888	0.1318	2.46	21.28	94.62
0.978	19.0	24947	6.4888	0.1328	2.46	21.28	94.62
0.9734	20.0	26260	6.3570	0.1328	2.46	21.28	94.62
0.9602	21.0	27573	6.2379	0.1335	2.46	21.28	94.62
0.9069	22.0	28886	6.3066	0.1334	2.46	21.28	94.62
0.8996	23.0	30199	6.2510	0.1328	2.46	21.28	94.62
0.893	24.0	31512	6.4094	0.1370	2.46	21.28	94.62
0.8494	25.0	32825	6.3232	0.1386	2.46	21.28	94.62
0.8507	26.0	34138	6.4262	0.1361	2.46	21.28	94.62
0.8065	27.0	35451	6.4156	0.1339	2.46	21.28	94.62
0.7956	28.0	36764	6.2878	0.1395	2.46	21.28	94.62
0.7889	29.0	38077	6.4994	0.1349	2.46	21.28	94.62
0.7645	30.0	39390	6.4939	0.1407	2.46	21.28	94.62
0.7548	31.0	40703	6.4849	0.1375	2.46	21.28	94.62
0.7494	32.0	42016	6.5542	0.1415	2.46	21.28	94.62
0.7162	33.0	43329	6.4573	0.1418	2.46	21.28	94.62
0.7109	34.0	44642	6.4910	0.1414	2.46	21.28	94.62
0.683	35.0	45955	6.4313	0.1411	2.46	21.28	94.62
0.6828	36.0	47268	6.3059	0.1456	2.46	21.28	94.62
0.6772	37.0	48581	6.3764	0.1464	2.46	21.28	94.62
0.652	38.0	49894	6.3437	0.1502	2.46	21.28	94.62
0.6533	39.0	51207	6.3493	0.1470	2.46	21.28	94.62
0.6527	40.0	52520	6.3078	0.1481	2.46	21.28	94.62
0.633	41.0	53833	6.5351	0.1413	2.46	21.28	94.62
0.6219	42.0	55146	6.3772	0.1492	2.46	21.28	94.62
0.6053	43.0	56459	6.4808	0.1481	2.46	21.28	94.62
0.5996	44.0	57772	6.5651	0.1480	2.46	21.28	94.62
0.5974	45.0	59085	6.5338	0.1488	2.46	21.28	94.62
0.5818	46.0	60398	6.3044	0.1524	2.46	21.28	94.62
0.5803	47.0	61711	6.5366	0.1514	2.46	21.28	94.62
0.573	48.0	63024	6.4783	0.1528	2.46	21.28	94.62
0.551	49.0	64337	6.4941	0.1540	2.46	21.28	94.62
0.5447	50.0	65650	6.4514	0.1528	2.46	21.28	94.62
0.5326	51.0	66963	6.3732	0.1547	2.46	21.28	94.62
0.5307	52.0	68276	6.5803	0.1546	2.46	21.28	94.62
0.5265	53.0	69589	6.2254	0.1594	2.46	21.28	94.62
0.5216	54.0	70902	6.2881	0.1574	2.46	21.28	94.62
0.5214	55.0	72215	6.4118	0.1564	2.46	21.28	94.62
0.5163	56.0	73528	6.4703	0.1574	2.46	21.28	94.62
0.4954	57.0	74841	6.3910	0.1602	2.46	21.28	94.62
0.4946	58.0	76154	6.4567	0.1607	2.46	21.28	94.62
0.4764	59.0	77467	6.4750	0.1592	2.46	21.28	94.62
0.4797	60.0	78780	6.5071	0.1580	2.46	21.28	94.62
0.4773	61.0	80093	6.2996	0.1649	2.46	21.28	94.62
0.4638	62.0	81406	6.3757	0.1582	2.46	21.35	94.62
0.4634	63.0	82719	6.4944	0.1579	2.46	21.35	94.62
0.4605	64.0	84032	6.6361	0.1573	2.46	21.35	94.62
0.4541	65.0	85345	6.5321	0.1566	2.46	21.35	94.62
0.447	66.0	86658	6.2949	0.1647	2.46	21.35	94.62
0.4392	67.0	87971	6.4294	0.1616	2.46	21.35	94.62
0.4319	68.0	89284	6.4686	0.1657	2.46	21.35	94.62
0.4321	69.0	90597	6.5044	0.1654	2.46	21.35	94.62
0.4239	70.0	91910	6.2884	0.1670	2.46	21.35	94.62
0.424	71.0	93223	6.4557	0.1650	2.46	21.35	94.62
0.4189	72.0	94536	6.5151	0.1643	2.46	21.35	94.62
0.4056	73.0	95849	6.4498	0.1685	2.46	21.35	94.62
0.4113	74.0	97162	6.4636	0.1672	2.46	21.35	94.62
0.4031	75.0	98475	6.6464	0.1627	2.46	21.35	94.62
0.3965	76.0	99788	6.5633	0.1686	2.46	21.35	94.62
0.393	77.0	101101	6.6878	0.1633	2.46	21.35	94.62
0.3958	78.0	102414	6.4100	0.1742	2.46	21.35	94.62
0.3848	79.0	103727	6.5372	0.1708	2.46	21.35	94.62
0.3785	80.0	105040	6.4460	0.1702	2.46	21.35	94.62
0.3709	81.0	106353	6.4497	0.1763	2.46	21.35	94.62
0.3692	82.0	107666	6.4494	0.1746	2.46	21.35	94.62
0.3667	83.0	108979	6.4787	0.1733	2.46	21.35	94.62
0.3642	84.0	110292	6.3792	0.1762	2.46	21.35	94.62
0.3648	85.0	111605	6.4105	0.1784	2.46	21.35	94.62
0.3595	86.0	112918	6.6821	0.1718	2.46	21.35	94.62
0.3575	87.0	114231	6.5187	0.1763	2.46	21.35	94.62
0.3512	88.0	115544	6.5861	0.1752	2.46	21.35	94.62
0.3416	89.0	116857	6.5337	0.1772	2.46	21.35	94.62
0.3454	90.0	118170	6.6075	0.1758	2.46	21.35	94.62
0.3401	91.0	119483	6.5369	0.1759	2.46	21.35	94.62
0.3361	92.0	120796	6.6148	0.1777	2.46	21.35	94.62
0.3377	93.0	122109	6.4843	0.1799	2.46	21.35	94.62
0.3344	94.0	123422	6.4471	0.1790	2.46	21.35	94.62
0.3262	95.0	124735	6.4506	0.1810	2.46	21.35	94.62
0.3228	96.0	126048	6.5665	0.1794	2.46	21.35	94.62
0.327	97.0	127361	6.5349	0.1793	2.46	21.35	94.62
0.3275	98.0	128674	6.5128	0.1799	2.46	21.35	94.62
0.321	99.0	129987	6.5574	0.1801	2.46	21.35	94.62
0.3217	100.0	131300	6.5510	0.1798	2.46	21.35	94.62

Framework versions

• Transformers 4.45.2
• Pytorch 2.6.0+hpu_1.20.0-543.git4952fce
• Datasets 3.5.0
• Tokenizers 0.20.3