MAE-CT-CPC-Dicotomized-v8-n0-m1

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

• Loss: 0.4681
• Accuracy: 0.7907

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: 1e-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
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 3500

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6769	0.02	70	0.6814	0.5938
0.7223	1.02	140	0.6953	0.5938
0.6628	2.02	210	0.6335	0.625
0.5096	3.02	280	0.6514	0.625
0.4739	4.02	350	0.6358	0.6562
0.4554	5.02	420	0.6272	0.6562
0.4818	6.02	490	0.7727	0.5938
0.4129	7.02	560	0.8222	0.6875
0.6301	8.02	630	0.8041	0.625
0.3809	9.02	700	0.8721	0.625
0.8071	10.02	770	1.1092	0.625
0.1888	11.02	840	1.1556	0.625
0.3762	12.02	910	1.3499	0.625
0.3502	13.02	980	1.5333	0.6562
0.1027	14.02	1050	1.6249	0.6562
0.177	15.02	1120	1.3758	0.6562
0.0998	16.02	1190	1.9514	0.6562
0.1749	17.02	1260	1.9120	0.6562
0.0145	18.02	1330	2.1036	0.625
0.0038	19.02	1400	2.0288	0.625
0.1262	20.02	1470	2.0193	0.6875
0.0203	21.02	1540	2.1937	0.6562
0.0002	22.02	1610	2.2922	0.625
0.0017	23.02	1680	2.1569	0.6562
0.0049	24.02	1750	2.2573	0.625
0.0231	25.02	1820	2.1460	0.6875
0.0001	26.02	1890	2.3566	0.6562
0.0001	27.02	1960	2.3822	0.5938
0.0001	28.02	2030	2.3178	0.6562
0.0004	29.02	2100	2.5492	0.625
0.0003	30.02	2170	2.7648	0.625
0.0001	31.02	2240	2.3949	0.625
0.0001	32.02	2310	2.4107	0.6562
0.0001	33.02	2380	2.6099	0.5938
0.0001	34.02	2450	2.8574	0.5625
0.0001	35.02	2520	2.5808	0.5938
0.0001	36.02	2590	2.6246	0.5938
0.0001	37.02	2660	2.7051	0.5938
0.0001	38.02	2730	2.5046	0.5938
0.0001	39.02	2800	2.5003	0.5938
0.0	40.02	2870	2.5460	0.625
0.0	41.02	2940	2.5397	0.625
0.0	42.02	3010	2.5384	0.625
0.0	43.02	3080	2.4849	0.625
0.0	44.02	3150	2.5847	0.6562
0.0	45.02	3220	2.5829	0.6562
0.0	46.02	3290	2.5809	0.6562
0.0	47.02	3360	2.5756	0.625
0.0001	48.02	3430	2.4744	0.6562
0.0	49.02	3500	2.4720	0.6562

Framework versions

• Transformers 4.45.2
• Pytorch 2.0.1+cu117
• Datasets 3.0.1
• Tokenizers 0.20.0