metadata

license: apache-2.0
base_model: google/vit-large-patch16-224
tags:
  - image-classification
  - vision
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: vit-large-patch16-224-finetuned-galaxy10-decals
    results: []

vit-large-patch16-224-finetuned-galaxy10-decals

This model is a fine-tuned version of google/vit-large-patch16-224 on the matthieulel/galaxy10_decals dataset. It achieves the following results on the evaluation set:

Loss: 0.6404
Accuracy: 0.8551
Precision: 0.8525
Recall: 0.8551
F1: 0.8526

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: 0.0001
train_batch_size: 128
eval_batch_size: 128
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 512
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
1.0358	0.99	31	0.7598	0.7362	0.7425	0.7362	0.7266
0.6822	1.98	62	0.6136	0.7920	0.7972	0.7920	0.7899
0.6535	2.98	93	0.5416	0.8106	0.8140	0.8106	0.8062
0.5696	4.0	125	0.5305	0.8162	0.8195	0.8162	0.8140
0.5435	4.99	156	0.5555	0.8145	0.8242	0.8145	0.8161
0.4621	5.98	187	0.5075	0.8298	0.8344	0.8298	0.8254
0.4479	6.98	218	0.5118	0.8281	0.8291	0.8281	0.8269
0.4318	8.0	250	0.5164	0.8196	0.8255	0.8196	0.8166
0.4011	8.99	281	0.5087	0.8410	0.8369	0.8410	0.8362
0.355	9.98	312	0.5063	0.8410	0.8433	0.8410	0.8405
0.3655	10.98	343	0.5419	0.8326	0.8343	0.8326	0.8305
0.3292	12.0	375	0.5134	0.8439	0.8442	0.8439	0.8415
0.3207	12.99	406	0.6285	0.8185	0.8293	0.8185	0.8178
0.2931	13.98	437	0.5627	0.8382	0.8395	0.8382	0.8371
0.2817	14.98	468	0.6059	0.8207	0.8278	0.8207	0.8215
0.2713	16.0	500	0.6140	0.8382	0.8367	0.8382	0.8337
0.233	16.99	531	0.5992	0.8382	0.8384	0.8382	0.8374
0.2313	17.98	562	0.6679	0.8292	0.8343	0.8292	0.8278
0.223	18.98	593	0.6501	0.8343	0.8386	0.8343	0.8347
0.2126	20.0	625	0.6731	0.8343	0.8304	0.8343	0.8296
0.2078	20.99	656	0.6335	0.8388	0.8410	0.8388	0.8383
0.201	21.98	687	0.6120	0.8506	0.8478	0.8506	0.8485
0.2045	22.98	718	0.6590	0.8410	0.8389	0.8410	0.8371
0.1759	24.0	750	0.6478	0.8489	0.8464	0.8489	0.8457
0.1856	24.99	781	0.6604	0.8444	0.8413	0.8444	0.8420
0.1766	25.98	812	0.6922	0.8501	0.8491	0.8501	0.8484
0.1841	26.98	843	0.6485	0.8501	0.8493	0.8501	0.8486
0.1707	28.0	875	0.6393	0.8467	0.8440	0.8467	0.8446
0.1792	28.99	906	0.6404	0.8551	0.8525	0.8551	0.8526
0.1713	29.76	930	0.6398	0.8534	0.8513	0.8534	0.8511

Framework versions

Transformers 4.37.2
Pytorch 2.3.0
Datasets 2.19.1
Tokenizers 0.15.1