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: 0.3575
Accuracy: 0.9456
Precision: 0.9498
Recall: 0.9456
F1 Score: 0.9473

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: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 128
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	0.94	4	0.3212	0.8475	0.8711	0.8475	0.7915
No log	1.88	8	0.2355	0.8983	0.8925	0.8983	0.8937
No log	2.82	12	0.3134	0.8644	0.8834	0.8644	0.8243
0.2493	4.0	17	0.2434	0.8814	0.8962	0.8814	0.8534
0.2493	4.94	21	0.3406	0.8983	0.9094	0.8983	0.8794
0.2493	5.88	25	0.1131	0.9322	0.9300	0.9322	0.9291
0.2493	6.82	29	0.1727	0.9153	0.9435	0.9153	0.9215
0.0374	8.0	34	0.6181	0.8644	0.8834	0.8644	0.8243
0.0374	8.94	38	0.3249	0.9153	0.9125	0.9153	0.9135
0.0374	9.88	42	0.5308	0.8983	0.8934	0.8983	0.8876
0.007	10.82	46	0.4767	0.9153	0.9119	0.9153	0.9090
0.007	12.0	51	0.3883	0.8983	0.8925	0.8983	0.8937
0.007	12.94	55	0.3627	0.8983	0.8934	0.8983	0.8876
0.007	13.88	59	0.2783	0.9492	0.9479	0.9492	0.9481
0.0012	14.82	63	0.1934	0.9492	0.9519	0.9492	0.9501
0.0012	16.0	68	0.1670	0.9661	0.9661	0.9661	0.9661
0.0012	16.94	72	0.1783	0.9492	0.9479	0.9492	0.9481
0.0001	17.88	76	0.4825	0.9322	0.9373	0.9322	0.9251
0.0001	18.82	80	0.9010	0.8983	0.9094	0.8983	0.8794
0.0001	20.0	85	0.1802	0.9661	0.9718	0.9661	0.9673
0.0001	20.94	89	0.5658	0.9153	0.9119	0.9153	0.9090
0.0037	21.88	93	0.8331	0.9322	0.9373	0.9322	0.9251
0.0037	22.82	97	0.8074	0.9153	0.9119	0.9153	0.9090
0.0037	24.0	102	0.4763	0.8814	0.8771	0.8814	0.8788
0.0002	24.94	106	0.5553	0.9153	0.9119	0.9153	0.9090
0.0002	25.88	110	0.8220	0.9153	0.9231	0.9153	0.9032
0.0002	26.82	114	0.5367	0.9322	0.9373	0.9322	0.9251
0.0002	28.0	119	0.4401	0.9153	0.9298	0.9153	0.9194
0.0037	28.94	123	0.4138	0.9153	0.9125	0.9153	0.9135
0.0037	29.88	127	0.7232	0.8983	0.9094	0.8983	0.8794
0.0037	30.82	131	0.3690	0.9322	0.9373	0.9322	0.9251
0.0115	32.0	136	0.2730	0.9322	0.9400	0.9322	0.9346
0.0115	32.94	140	0.2101	0.9661	0.9661	0.9661	0.9661
0.0115	33.88	144	0.1814	0.9661	0.9661	0.9661	0.9661
0.0115	34.82	148	0.1641	0.9661	0.9661	0.9661	0.9661
0.0013	36.0	153	0.1600	0.9492	0.9479	0.9492	0.9481
0.0013	36.94	157	0.1709	0.9661	0.9674	0.9661	0.9646
0.0013	37.88	161	0.1913	0.9661	0.9674	0.9661	0.9646
0.0001	38.82	165	0.2047	0.9661	0.9674	0.9661	0.9646
0.0001	40.0	170	0.2030	0.9661	0.9674	0.9661	0.9646
0.0001	40.94	174	0.1960	0.9661	0.9674	0.9661	0.9646
0.0001	41.88	178	0.1936	0.9661	0.9674	0.9661	0.9646
0.0003	42.35	180	0.1934	0.9661	0.9674	0.9661	0.9646

Framework versions

Transformers 4.38.2
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.15.2