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: 0.2254
• Accuracy: 0.9180

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: 32
• eval_batch_size: 32
• 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: 325

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.0123	4	0.6042	0.8689
No log	1.0123	8	0.5912	0.8852
0.1557	2.0123	12	0.5992	0.8852
0.1557	3.0123	16	0.5476	0.8852
0.1353	4.0123	20	0.5838	0.8689
0.1353	5.0123	24	0.5512	0.8689
0.1353	6.0123	28	0.5185	0.8689
0.1002	7.0123	32	0.4810	0.9016
0.1002	8.0123	36	0.3758	0.9016
0.0612	9.0123	40	0.4661	0.8852
0.0612	10.0123	44	0.4693	0.8525
0.0612	11.0123	48	0.3250	0.8689
0.0395	12.0123	52	0.3807	0.8852
0.0395	13.0123	56	0.3268	0.9016
0.0294	14.0123	60	0.3572	0.8689
0.0294	15.0123	64	0.4752	0.8852
0.0294	16.0123	68	0.3392	0.8852
0.0232	17.0123	72	0.5777	0.8689
0.0232	18.0123	76	0.3572	0.9016
0.0168	19.0123	80	0.6077	0.8361
0.0168	20.0123	84	0.3536	0.9016
0.0168	21.0123	88	0.4749	0.8852
0.0141	22.0123	92	0.3578	0.9016
0.0141	23.0123	96	0.3149	0.8852
0.0113	24.0123	100	0.3504	0.8852
0.0113	25.0123	104	0.2214	0.9180
0.0113	26.0123	108	0.2442	0.9180
0.0143	27.0123	112	0.4498	0.8852
0.0143	28.0123	116	0.3842	0.8689
0.0104	29.0123	120	0.3404	0.8689
0.0104	30.0123	124	0.2934	0.9016
0.0104	31.0123	128	0.2977	0.9180
0.009	32.0123	132	0.3295	0.8852
0.009	33.0123	136	0.3387	0.8689
0.0074	34.0123	140	0.3298	0.8852
0.0074	35.0123	144	0.3200	0.8852
0.0074	36.0123	148	0.3098	0.9016
0.0069	37.0123	152	0.3426	0.8852
0.0069	38.0123	156	0.3258	0.9016
0.0063	39.0123	160	0.3417	0.8852
0.0063	40.0123	164	0.3744	0.8852
0.0063	41.0123	168	0.3680	0.8852
0.0058	42.0123	172	0.3940	0.8689
0.0058	43.0123	176	0.3287	0.8852
0.0163	44.0123	180	0.2990	0.8852
0.0163	45.0123	184	0.3104	0.8852
0.0163	46.0123	188	0.3738	0.8689
0.0059	47.0123	192	0.3384	0.8852
0.0059	48.0123	196	0.3533	0.8852
0.0058	49.0123	200	0.3802	0.9016
0.0058	50.0123	204	0.4168	0.9016
0.0058	51.0123	208	0.3361	0.8852
0.0053	52.0123	212	0.3255	0.8852
0.0053	53.0123	216	0.3412	0.8852
0.0057	54.0123	220	0.3386	0.8852
0.0057	55.0123	224	0.3742	0.8689
0.0057	56.0123	228	0.3378	0.8689
0.0051	57.0123	232	0.3133	0.8852
0.0051	58.0123	236	0.2935	0.8689
0.0048	59.0123	240	0.2900	0.8689
0.0048	60.0123	244	0.2831	0.8852
0.0048	61.0123	248	0.3001	0.8852
0.0048	62.0123	252	0.3225	0.8852
0.0048	63.0123	256	0.3803	0.8852
0.0046	64.0123	260	0.3333	0.8852
0.0046	65.0123	264	0.3481	0.8852
0.0046	66.0123	268	0.3178	0.8852
0.0045	67.0123	272	0.3117	0.8852
0.0045	68.0123	276	0.3241	0.8852
0.0044	69.0123	280	0.3243	0.8852
0.0044	70.0123	284	0.3431	0.8852
0.0044	71.0123	288	0.3221	0.8852
0.0044	72.0123	292	0.2979	0.8852
0.0044	73.0123	296	0.3110	0.8852
0.0043	74.0123	300	0.3329	0.8852
0.0043	75.0123	304	0.3023	0.8852
0.0043	76.0123	308	0.3122	0.8852
0.0045	77.0123	312	0.3068	0.8852
0.0045	78.0123	316	0.3510	0.8852
0.0043	79.0123	320	0.3064	0.8852
0.0043	80.0123	324	0.3018	0.8852
0.0043	81.0031	325	0.2982	0.8852

Framework versions

• Transformers 4.48.0
• Pytorch 2.0.1+cu118
• Datasets 3.2.0
• Tokenizers 0.21.0