CTMAE-P2-V5-3g-S5

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: 0.5090
• Accuracy: 0.8667

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: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 13050

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.8752	0.02	261	2.0009	0.4667
0.5365	1.02	522	2.8415	0.4667
1.8181	2.02	783	2.1331	0.4667
0.6674	3.02	1044	2.5703	0.4667
1.7586	4.02	1305	1.1413	0.4667
1.0731	5.02	1566	1.8862	0.4667
1.3539	6.02	1827	1.8277	0.4667
0.8906	7.02	2088	1.7559	0.4667
1.5706	8.02	2349	2.0014	0.4667
0.3113	9.02	2610	1.3058	0.6667
1.6658	10.02	2871	1.5835	0.6444
1.3587	11.02	3132	0.8309	0.7556
0.31	12.02	3393	0.8154	0.7778
1.2834	13.02	3654	0.5090	0.8667
0.7111	14.02	3915	1.2500	0.6889
2.3551	15.02	4176	0.6881	0.8222
0.2734	16.02	4437	0.4506	0.8444
0.9675	17.02	4698	1.7515	0.6667
0.449	18.02	4959	0.7240	0.7778
0.5843	19.02	5220	0.9561	0.7778
1.0949	20.02	5481	1.2866	0.6889
1.2073	21.02	5742	1.0336	0.7556
1.3534	22.02	6003	1.8029	0.7111
0.0423	23.02	6264	1.4571	0.7111
1.0068	24.02	6525	1.7790	0.6444
1.5772	25.02	6786	1.7893	0.6667
0.8409	26.02	7047	1.6454	0.6667
0.6828	27.02	7308	1.8521	0.6889
0.5191	28.02	7569	1.2734	0.7556
0.4537	29.02	7830	1.8099	0.7111
0.003	30.02	8091	1.5860	0.7333
0.0004	31.02	8352	2.2568	0.6444
0.1452	32.02	8613	2.4112	0.6444
0.3815	33.02	8874	1.3679	0.7556
0.0013	34.02	9135	1.8306	0.7111
0.7655	35.02	9396	1.4608	0.7333
0.003	36.02	9657	2.2029	0.6667
0.0246	37.02	9918	2.7586	0.6222
0.0007	38.02	10179	2.6804	0.6444
0.5967	39.02	10440	2.5969	0.6444
0.0006	40.02	10701	2.6381	0.6444
0.0004	41.02	10962	2.9591	0.6222
0.0004	42.02	11223	2.1240	0.7111
1.109	43.02	11484	2.7634	0.6
1.074	44.02	11745	2.1595	0.7111
0.0016	45.02	12006	1.8538	0.7556
0.0002	46.02	12267	1.9419	0.7556
0.0005	47.02	12528	1.8378	0.7778
0.0001	48.02	12789	2.0625	0.7333
0.59	49.02	13050	2.0340	0.7333

Framework versions

• Transformers 4.46.2
• Pytorch 2.0.1+cu117
• Datasets 3.0.1
• Tokenizers 0.20.0