metadata

license: apache-2.0
base_model: google/vit-base-patch32-224-in21k
tags:
  - image-classification
  - vision
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: vit-base-patch32-224-in21k-finetuned-galaxy10-decals
    results: []

vit-base-patch32-224-in21k-finetuned-galaxy10-decals

This model is a fine-tuned version of google/vit-base-patch32-224-in21k on the matthieulel/galaxy10_decals dataset. It achieves the following results on the evaluation set:

Loss: 0.5180
Accuracy: 0.8382
Precision: 0.8363
Recall: 0.8382
F1: 0.8346

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: 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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.4731	0.99	124	1.3850	0.6110	0.5791	0.6110	0.5797
0.9858	2.0	249	0.8900	0.7508	0.7578	0.7508	0.7337
0.9475	3.0	374	0.7799	0.7599	0.7667	0.7599	0.7559
0.7778	4.0	499	0.6798	0.7779	0.7825	0.7779	0.7729
0.6831	4.99	623	0.6352	0.7914	0.7916	0.7914	0.7889
0.6953	6.0	748	0.5931	0.8044	0.8076	0.8044	0.8023
0.6725	7.0	873	0.7304	0.7537	0.7671	0.7537	0.7519
0.5648	8.0	998	0.6352	0.7909	0.7961	0.7909	0.7868
0.6127	8.99	1122	0.6087	0.7858	0.7879	0.7858	0.7820
0.529	10.0	1247	0.5827	0.8072	0.8074	0.8072	0.8041
0.5212	11.0	1372	0.5787	0.8179	0.8177	0.8179	0.8108
0.4665	12.0	1497	0.5597	0.8168	0.8213	0.8168	0.8134
0.5123	12.99	1621	0.5840	0.8044	0.8163	0.8044	0.8044
0.4918	14.0	1746	0.5592	0.8219	0.8221	0.8219	0.8195
0.4733	15.0	1871	0.5180	0.8382	0.8363	0.8382	0.8346
0.4552	16.0	1996	0.5673	0.8174	0.8181	0.8174	0.8153
0.4004	16.99	2120	0.5711	0.8224	0.8239	0.8224	0.8199
0.3359	18.0	2245	0.5813	0.8168	0.8153	0.8168	0.8147
0.4069	19.0	2370	0.5482	0.8343	0.8352	0.8343	0.8307
0.3783	20.0	2495	0.5658	0.8179	0.8169	0.8179	0.8150
0.3293	20.99	2619	0.5647	0.8247	0.8234	0.8247	0.8230
0.3214	22.0	2744	0.5654	0.8309	0.8289	0.8309	0.8293
0.3285	23.0	2869	0.5943	0.8213	0.8226	0.8213	0.8201
0.2934	24.0	2994	0.5931	0.8264	0.8287	0.8264	0.8259
0.3051	24.99	3118	0.5788	0.8309	0.8325	0.8309	0.8303
0.2911	26.0	3243	0.5700	0.8377	0.8354	0.8377	0.8358
0.2893	27.0	3368	0.5971	0.8286	0.8320	0.8286	0.8291
0.2794	28.0	3493	0.5908	0.8315	0.8307	0.8315	0.8303
0.2506	28.99	3617	0.5914	0.8309	0.8314	0.8309	0.8306
0.2421	29.82	3720	0.5861	0.8365	0.8366	0.8365	0.8359

Framework versions

Transformers 4.37.2
Pytorch 2.3.0
Datasets 2.19.1
Tokenizers 0.15.1