videomae-base-finetuned-bleeding

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: 3.4322
• Accuracy: 0.5

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: 2
• eval_batch_size: 2
• 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: 250

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.024	6	0.7224	0.75
0.7131	1.024	12	0.7031	0.5
0.7131	2.024	18	0.6849	0.75
0.6901	3.024	24	0.6763	0.5
0.6385	4.024	30	0.6848	0.5
0.6385	5.024	36	0.6625	0.5
0.6959	6.024	42	0.9288	0.5
0.6959	7.024	48	0.8589	0.5
0.5537	8.024	54	1.2295	0.5
0.5147	9.024	60	1.8338	0.5
0.5147	10.024	66	2.1135	0.5
0.9219	11.024	72	2.0003	0.5
0.9219	12.024	78	2.4492	0.5
1.087	13.024	84	2.6106	0.5
0.5866	14.024	90	2.9419	0.5
0.5866	15.024	96	2.4348	0.5
0.6801	16.024	102	2.6519	0.5
0.6801	17.024	108	2.6595	0.5
0.216	18.024	114	2.5805	0.5
1.1218	19.024	120	2.0237	0.5
1.1218	20.024	126	2.1719	0.5
0.1052	21.024	132	2.4262	0.5
0.1052	22.024	138	2.1011	0.5
0.0107	23.024	144	2.4142	0.5
0.1062	24.024	150	3.1635	0.5
0.1062	25.024	156	3.3490	0.5
0.0014	26.024	162	3.5096	0.5
0.0014	27.024	168	3.1308	0.5
0.6518	28.024	174	2.9847	0.5
0.4084	29.024	180	3.2943	0.5
0.4084	30.024	186	3.3435	0.5
0.3778	31.024	192	3.4386	0.5
0.3778	32.024	198	3.5861	0.5
0.1195	33.024	204	3.5148	0.5
0.2964	34.024	210	3.3186	0.5
0.2964	35.024	216	3.3009	0.5
0.048	36.024	222	3.3014	0.5
0.048	37.024	228	3.2850	0.5
0.0009	38.024	234	3.2852	0.5
0.0047	39.024	240	3.3801	0.5
0.0047	40.024	246	3.4257	0.5
0.0008	41.016	250	3.4322	0.5

Framework versions

• Transformers 4.40.1
• Pytorch 1.12.0
• Datasets 2.19.1
• Tokenizers 0.19.1