MAE-CT-M1N0-v10

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: 0.7797
• Accuracy: 0.6377

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: 3000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6946	0.02	60	0.6537	0.6571
0.6507	1.02	120	0.6382	0.6571
0.5703	2.02	180	0.6546	0.6571
0.5151	3.02	240	0.6290	0.6571
0.4583	4.02	300	0.5577	0.6571
0.4865	5.02	360	0.4877	0.7714
0.5124	6.02	420	0.5929	0.6571
0.3922	7.02	480	0.8184	0.7429
0.4062	8.02	540	1.0667	0.7429
0.4161	9.02	600	1.6347	0.5714
0.194	10.02	660	1.5608	0.6286
0.1609	11.02	720	1.8647	0.5714
0.1766	12.02	780	2.0934	0.5714
0.1442	13.02	840	2.0064	0.6
0.1134	14.02	900	1.7492	0.6286
0.0021	15.02	960	2.0751	0.6571
0.1016	16.02	1020	1.7400	0.6571
0.0056	17.02	1080	2.6220	0.6
0.0206	18.02	1140	2.5699	0.6
0.0003	19.02	1200	1.8053	0.7429
0.0219	20.02	1260	1.8495	0.6857
0.0347	21.02	1320	3.1257	0.5429
0.0002	22.02	1380	1.4274	0.7714
0.1377	23.02	1440	2.8967	0.5714
0.0001	24.02	1500	2.7693	0.5714
0.0007	25.02	1560	2.7892	0.5714
0.135	26.02	1620	3.0140	0.5429
0.0002	27.02	1680	2.7728	0.6
0.0003	28.02	1740	2.3857	0.6
0.0001	29.02	1800	3.2607	0.5714
0.0001	30.02	1860	2.8166	0.6
0.0001	31.02	1920	2.8641	0.6
0.0001	32.02	1980	3.1273	0.5714
0.0001	33.02	2040	2.6391	0.6286
0.0001	34.02	2100	2.8301	0.6286
0.0001	35.02	2160	2.9671	0.6
0.0013	36.02	2220	2.7573	0.6286
0.0001	37.02	2280	3.0395	0.5714
0.0001	38.02	2340	2.9163	0.6286
0.0001	39.02	2400	2.9096	0.6
0.2136	40.02	2460	3.1780	0.5714
0.0001	41.02	2520	3.1003	0.5714
0.0001	42.02	2580	3.1442	0.5714
0.0001	43.02	2640	3.2157	0.5714
0.0001	44.02	2700	3.1772	0.5714
0.0001	45.02	2760	3.1787	0.5714
0.0	46.02	2820	3.1770	0.5714
0.0001	47.02	2880	3.1718	0.5714
0.0001	48.02	2940	3.1726	0.5714
0.0001	49.02	3000	3.1944	0.5714

Framework versions

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