WS800_BEiT_42895082

This model is a fine-tuned version of microsoft/beit-base-patch16-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0845
• Accuracy: 0.975

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	5	0.6815	0.9375
No log	2.0	10	0.6041	0.95
No log	3.0	15	0.4946	0.9125
No log	4.0	20	0.3233	0.975
No log	5.0	25	0.2158	0.9875
No log	6.0	30	0.1514	0.9875
No log	7.0	35	0.1109	0.9875
No log	8.0	40	0.0909	0.9875
No log	9.0	45	0.0886	0.975
0.3029	10.0	50	0.1020	0.975
0.3029	11.0	55	0.1155	0.975
0.3029	12.0	60	0.1197	0.975
0.3029	13.0	65	0.1247	0.975
0.3029	14.0	70	0.1021	0.975
0.3029	15.0	75	0.2528	0.95
0.3029	16.0	80	0.1866	0.9625
0.3029	17.0	85	0.2019	0.9625
0.3029	18.0	90	0.1456	0.9625
0.3029	19.0	95	0.1014	0.975
0.0143	20.0	100	0.0845	0.975

Framework versions

• Transformers 4.36.2
• Pytorch 2.1.2+cu118
• Datasets 2.16.1
• Tokenizers 0.15.0