videomae-base-face

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: 1.8802
• Accuracy: 0.6944

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: 5
• eval_batch_size: 5
• 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: 1200

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.6771	16.0033	100	1.5844	0.2222
0.2052	33.0017	200	1.8833	0.4722
0.6001	49.005	300	1.4486	0.6944
0.3118	66.0033	400	0.1618	0.9722
0.0046	83.0017	500	2.1274	0.6944
0.0528	99.005	600	1.8246	0.7222
0.0174	116.0033	700	1.9694	0.7222
0.2597	133.0017	800	2.0549	0.6944
0.0505	149.005	900	1.9087	0.7222
0.0014	166.0033	1000	2.0244	0.6944
0.0324	183.0017	1100	1.5456	0.75
0.0011	199.005	1200	1.8802	0.6944

Framework versions

• Transformers 4.45.0
• Pytorch 2.4.1+cu118
• Datasets 3.0.0
• Tokenizers 0.20.0