This model is a fine-tuned version of microsoft/beit-base-patch16-224-pt22k on the matthieulel/galaxy10_decals dataset. It achieves the following results on the evaluation set:

Loss: 0.4135
Accuracy: 0.8653
Precision: 0.8629
Recall: 0.8653
F1: 0.8629

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: 64
eval_batch_size: 64
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 256
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.7405	0.99	62	1.4465	0.4634	0.4965	0.4634	0.4178
1.0688	2.0	125	0.8733	0.7108	0.7043	0.7108	0.6971
0.8939	2.99	187	0.7320	0.7373	0.7472	0.7373	0.7331
0.7122	4.0	250	0.6952	0.7700	0.7632	0.7700	0.7596
0.6654	4.99	312	0.5980	0.7914	0.7917	0.7914	0.7910
0.667	6.0	375	0.6021	0.7892	0.7996	0.7892	0.7822
0.615	6.99	437	0.5403	0.8083	0.8098	0.8083	0.8019
0.5685	8.0	500	0.4871	0.8230	0.8203	0.8230	0.8200
0.5764	8.99	562	0.4781	0.8315	0.8332	0.8315	0.8304
0.5372	10.0	625	0.4768	0.8331	0.8327	0.8331	0.8285
0.5361	10.99	687	0.4883	0.8354	0.8368	0.8354	0.8338
0.5051	12.0	750	0.4693	0.8354	0.8338	0.8354	0.8328
0.4938	12.99	812	0.4517	0.8416	0.8418	0.8416	0.8396
0.4883	14.0	875	0.4434	0.8472	0.8487	0.8472	0.8442
0.4544	14.99	937	0.4522	0.8427	0.8411	0.8427	0.8395
0.4352	16.0	1000	0.4369	0.8512	0.8493	0.8512	0.8488
0.4107	16.99	1062	0.4545	0.8540	0.8545	0.8540	0.8504
0.4354	18.0	1125	0.4304	0.8563	0.8553	0.8563	0.8550
0.4337	18.99	1187	0.4293	0.8579	0.8572	0.8579	0.8564
0.4252	20.0	1250	0.4329	0.8506	0.8535	0.8506	0.8473
0.3923	20.99	1312	0.4171	0.8602	0.8602	0.8602	0.8586
0.4216	22.0	1375	0.4191	0.8568	0.8574	0.8568	0.8548
0.3847	22.99	1437	0.4378	0.8517	0.8509	0.8517	0.8509
0.3606	24.0	1500	0.4403	0.8585	0.8571	0.8585	0.8567
0.3739	24.99	1562	0.4228	0.8619	0.8598	0.8619	0.8594
0.3291	26.0	1625	0.4112	0.8602	0.8579	0.8602	0.8587
0.3441	26.99	1687	0.4214	0.8608	0.8592	0.8608	0.8588
0.3649	28.0	1750	0.4135	0.8653	0.8629	0.8653	0.8629
0.345	28.99	1812	0.4177	0.8608	0.8594	0.8608	0.8589
0.3435	29.76	1860	0.4188	0.8625	0.8610	0.8625	0.8606

Framework versions

Transformers 4.37.2
Pytorch 2.3.0
Datasets 2.19.1
Tokenizers 0.15.1

matthieulel/beit-base-patch16-224-pt22k-finetuned-galaxy10-decals

beit-base-patch16-224-pt22k-finetuned-galaxy10-decals

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

matthieulel/
beit-base-patch16-224-pt22k-finetuned-galaxy10-decals