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: 0.7149
• Accuracy: 0.9038

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: 1
• eval_batch_size: 1
• 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: 15000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0174	0.02	300	0.2750	0.9423
0.0218	1.02	600	0.7020	0.8269
0.5348	2.02	900	1.1836	0.7692
0.9667	3.02	1200	1.9316	0.5962
2.0242	4.02	1500	1.5680	0.6923
0.7852	5.02	1800	0.5868	0.9038
1.9104	6.02	2100	1.8121	0.7115
1.1466	7.02	2400	1.3801	0.75
0.0025	8.02	2700	1.2799	0.7692
0.0005	9.02	3000	1.8073	0.7115
0.0005	10.02	3300	0.4820	0.9231
0.5816	11.02	3600	0.7625	0.8846
0.0014	12.02	3900	0.4762	0.9231
0.0694	13.02	4200	1.3250	0.8077
0.0002	14.02	4500	0.9637	0.8654
0.0003	15.02	4800	0.4808	0.9231
0.0003	16.02	5100	0.8623	0.8846
0.0002	17.02	5400	0.6881	0.9231
0.0004	18.02	5700	0.5577	0.9038
0.0001	19.02	6000	0.5069	0.9231
0.4994	20.02	6300	0.3667	0.9423
0.0002	21.02	6600	0.3666	0.9423
1.0279	22.02	6900	1.0781	0.8654
0.0135	23.02	7200	2.2670	0.7308
0.0002	24.02	7500	0.1732	0.9615
0.0002	25.02	7800	0.4422	0.9423
0.0001	26.02	8100	0.8196	0.8846
0.0001	27.02	8400	0.8037	0.8846
0.0001	28.02	8700	0.8696	0.8846
0.0002	29.02	9000	0.7887	0.9231
0.7745	30.02	9300	0.3868	0.9423
0.0001	31.02	9600	0.4386	0.9423
0.0002	32.02	9900	0.4036	0.9423
0.0001	33.02	10200	0.3513	0.9423
0.0001	34.02	10500	0.3075	0.9423
0.0001	35.02	10800	0.5712	0.9231
0.0005	36.02	11100	0.6482	0.9231
0.0001	37.02	11400	0.8843	0.9038
0.0001	38.02	11700	0.9147	0.8846
0.0001	39.02	12000	0.6891	0.9038
0.0001	40.02	12300	0.8976	0.8846
0.0001	41.02	12600	1.6405	0.8077
0.0001	42.02	12900	1.0550	0.8654
0.0	43.02	13200	1.0356	0.8654
0.0	44.02	13500	1.0037	0.8462
0.0	45.02	13800	0.9632	0.8654
0.0	46.02	14100	0.6649	0.9231
0.0	47.02	14400	0.8702	0.8846
0.0	48.02	14700	0.7201	0.9038
0.0	49.02	15000	0.7149	0.9038

Framework versions

• Transformers 4.40.0
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1