videomae-base-finetuned-ElderReact-anger-balanced-hp

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: 0.6938
• Accuracy: 0.4672

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: 0.001
• train_batch_size: 16
• eval_batch_size: 16
• 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: 480

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7532	0.05	25	0.7078	0.5238
0.7571	1.05	50	0.7034	0.4762
0.7357	2.05	75	0.7080	0.4429
0.6976	3.05	100	0.7160	0.5238
0.7131	4.05	125	0.6893	0.4714
0.7275	5.05	150	0.8350	0.4929
0.7334	6.05	175	0.7127	0.4738
0.7274	7.05	200	0.7088	0.5048
0.697	8.05	225	0.6911	0.5190
0.7605	9.05	250	0.7296	0.4976
0.7105	10.05	275	0.7100	0.4833
0.6745	11.05	300	0.7271	0.4548
0.7166	12.05	325	0.6955	0.5286
0.6849	13.05	350	0.6981	0.4976
0.6978	14.05	375	0.6976	0.4952
0.6928	15.05	400	0.6941	0.5405
0.7057	16.05	425	0.7022	0.5
0.6842	17.05	450	0.6943	0.4738
0.6824	18.05	475	0.6945	0.5167
0.7065	19.01	480	0.6948	0.5143

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2