metadata

license: apache-2.0
base_model: microsoft/beit-base-patch16-224
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
model-index:
  - name: beit-base-patch16-224
    results: []

beit-base-patch16-224

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: 0.8528
Accuracy: 0.8268
Precision: 0.8303
Recall: 0.8268
F1 Score: 0.8283

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: 48
eval_batch_size: 48
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 192
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 45

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1 Score
No log	0.8	2	0.6993	0.5882	0.5390	0.5882	0.5541
No log	2.0	5	0.5971	0.6863	0.6806	0.6863	0.6033
No log	2.8	7	0.5306	0.8039	0.8000	0.8039	0.8006
No log	4.0	10	0.4828	0.7255	0.7229	0.7255	0.6859
No log	4.8	12	0.3812	0.7843	0.7786	0.7843	0.7784
0.5413	6.0	15	0.5268	0.7451	0.7461	0.7451	0.7141
0.5413	6.8	17	0.5349	0.7451	0.8556	0.7451	0.7502
0.5413	8.0	20	0.4120	0.8039	0.8485	0.8039	0.7756
0.5413	8.8	22	0.3156	0.8039	0.8003	0.8039	0.7963
0.5413	10.0	25	0.3217	0.8039	0.8061	0.8039	0.7909
0.5413	10.8	27	0.5161	0.7843	0.7870	0.7843	0.7664
0.0919	12.0	30	0.3677	0.8431	0.8498	0.8431	0.8451
0.0919	12.8	32	0.4631	0.8431	0.8407	0.8431	0.8405
0.0919	14.0	35	0.5001	0.8235	0.8214	0.8235	0.8221
0.0919	14.8	37	0.4489	0.8431	0.8431	0.8431	0.8431
0.0919	16.0	40	0.5892	0.7843	0.7799	0.7843	0.7731
0.0919	16.8	42	0.6579	0.7843	0.7799	0.7843	0.7731
0.006	18.0	45	0.7038	0.7843	0.7799	0.7843	0.7731
0.006	18.8	47	0.5864	0.8627	0.8737	0.8627	0.8651
0.006	20.0	50	0.5488	0.8627	0.8737	0.8627	0.8651
0.006	20.8	52	0.6651	0.8039	0.8003	0.8039	0.7963
0.006	22.0	55	0.6265	0.8039	0.8000	0.8039	0.8006
0.006	22.8	57	0.5229	0.8627	0.8653	0.8627	0.8637
0.0048	24.0	60	0.5421	0.8627	0.8653	0.8627	0.8637
0.0048	24.8	62	0.6335	0.8235	0.8205	0.8235	0.8187
0.0048	26.0	65	1.0379	0.8039	0.8201	0.8039	0.7841
0.0048	26.8	67	0.9758	0.8235	0.8366	0.8235	0.8089
0.0048	28.0	70	0.6117	0.8235	0.8205	0.8235	0.8187
0.0048	28.8	72	0.5403	0.8627	0.8613	0.8627	0.8617
0.0063	30.0	75	0.6469	0.8431	0.8407	0.8431	0.8405
0.0063	30.8	77	0.7014	0.8235	0.8205	0.8235	0.8187
0.0063	32.0	80	0.7514	0.8235	0.8205	0.8235	0.8187
0.0063	32.8	82	0.7771	0.8235	0.8248	0.8235	0.8144
0.0063	34.0	85	0.7599	0.8039	0.8003	0.8039	0.7963
0.0063	34.8	87	0.7554	0.8039	0.8003	0.8039	0.7963
0.0045	36.0	90	0.7308	0.8039	0.8003	0.8039	0.7963

Framework versions

Transformers 4.40.1
Pytorch 2.2.1+cu121
Datasets 2.19.0
Tokenizers 0.19.1