videomae-base-finetuned-ucf101-subset

This model is a fine-tuned version of MCG-NJU/videomae-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1677
• Accuracy: 0.9806

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 1850

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.3124	0.0205	38	2.2579	0.1857
2.1472	1.0205	76	2.0229	0.4286
1.3692	2.0205	114	1.3557	0.4857
0.6782	3.0205	152	0.5930	0.7571
0.4244	4.0205	190	0.4616	0.8286
0.2646	5.0205	228	0.3399	0.9
0.1	6.0205	266	0.1133	0.9714
0.2128	7.0205	304	0.7523	0.8
0.0967	8.0205	342	0.5267	0.8429
0.0971	9.0205	380	0.3907	0.9143
0.0939	10.0205	418	0.1479	0.9571
0.2974	11.0205	456	0.3272	0.9
0.0306	12.0205	494	0.2917	0.9
0.0036	13.0205	532	0.1893	0.9429
0.0742	14.0205	570	0.2095	0.9429
0.0071	15.0205	608	0.1195	0.9714
0.0027	16.0205	646	0.1051	0.9571
0.0022	17.0205	684	0.0845	0.9714
0.0019	18.0205	722	0.2177	0.9286
0.0015	19.0205	760	0.2222	0.9571
0.01	20.0205	798	0.0353	0.9857
0.0045	21.0205	836	0.0630	0.9714
0.0014	22.0205	874	0.0316	0.9714
0.0014	23.0205	912	0.0420	0.9857
0.002	24.0205	950	0.3080	0.9286
0.1012	25.0205	988	0.1244	0.9571
0.0012	26.0205	1026	0.1970	0.9429
0.0011	27.0205	1064	0.1381	0.9571
0.0012	28.0205	1102	0.0515	0.9714
0.0022	29.0205	1140	0.1643	0.9571
0.0909	30.0205	1178	0.1082	0.9714
0.001	31.0205	1216	0.0079	1.0
0.001	32.0205	1254	0.0047	1.0
0.001	33.0205	1292	0.0038	1.0
0.001	34.0205	1330	0.0039	1.0
0.0013	35.0205	1368	0.3373	0.9286
0.0009	36.0205	1406	0.1716	0.9714
0.0011	37.0205	1444	0.1124	0.9857
0.0008	38.0205	1482	0.1068	0.9857
0.0008	39.0205	1520	0.0920	0.9857
0.0008	40.0205	1558	0.0893	0.9857
0.0008	41.0205	1596	0.0890	0.9857
0.0009	42.0205	1634	0.0912	0.9857
0.0008	43.0205	1672	0.0898	0.9857
0.0008	44.0205	1710	0.0868	0.9857
0.0008	45.0205	1748	0.0758	0.9857
0.0008	46.0205	1786	0.0743	0.9857
0.0117	47.0205	1824	0.0734	0.9857
0.0008	48.0141	1850	0.0738	0.9857

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.0