videomae-base-ASD_Eye_Contact_v1

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: 0.6104
• Accuracy: 0.8889

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: 8
• eval_batch_size: 8
• 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.1	1	0.7788	0.2326
No log	1.1	2	0.6849	0.5581
No log	2.1	3	0.6217	0.9767
No log	3.1	4	0.5667	0.9767
No log	4.1	5	0.5197	0.9767
No log	5.1	6	0.4813	0.9767
No log	6.1	7	0.4521	0.9767
No log	7.1	8	0.4296	1.0
No log	8.1	9	0.4158	1.0
0.5774	9.1	10	0.4091	1.0

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1