metadata

license: apache-2.0
base_model: microsoft/swinv2-base-patch4-window12to16-192to256-22kto1k-ft
tags:
  - image-classification
  - vision
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: >-
      swinv2-base-patch4-window12to16-192to256-22kto1k-ft-finetuned-galaxy10-decals
    results: []

swinv2-base-patch4-window12to16-192to256-22kto1k-ft-finetuned-galaxy10-decals

This model is a fine-tuned version of microsoft/swinv2-base-patch4-window12to16-192to256-22kto1k-ft on the matthieulel/galaxy10_decals dataset. It achieves the following results on the evaluation set:

Loss: 0.6138
Accuracy: 0.8653
Precision: 0.8633
Recall: 0.8653
F1: 0.8633

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
1.1028	0.99	62	0.8747	0.6815	0.7019	0.6815	0.6725
0.7637	2.0	125	0.6110	0.7993	0.8032	0.7993	0.7944
0.702	2.99	187	0.5407	0.8179	0.8282	0.8179	0.8201
0.6131	4.0	250	0.5038	0.8326	0.8356	0.8326	0.8276
0.5453	4.99	312	0.4523	0.8534	0.8547	0.8534	0.8528
0.5409	6.0	375	0.4908	0.8377	0.8389	0.8377	0.8339
0.5246	6.99	437	0.4583	0.8478	0.8509	0.8478	0.8486
0.478	8.0	500	0.4417	0.8506	0.8529	0.8506	0.8486
0.4845	8.99	562	0.4344	0.8596	0.8591	0.8596	0.8565
0.4228	10.0	625	0.4580	0.8478	0.8488	0.8478	0.8462
0.4414	10.99	687	0.4520	0.8534	0.8539	0.8534	0.8525
0.3783	12.0	750	0.4776	0.8517	0.8504	0.8517	0.8501
0.407	12.99	812	0.4800	0.8478	0.8482	0.8478	0.8444
0.3944	14.0	875	0.4541	0.8630	0.8639	0.8630	0.8618
0.3563	14.99	937	0.4848	0.8534	0.8531	0.8534	0.8523
0.3576	16.0	1000	0.4877	0.8540	0.8526	0.8540	0.8522
0.317	16.99	1062	0.5122	0.8551	0.8572	0.8551	0.8546
0.3439	18.0	1125	0.5073	0.8484	0.8509	0.8484	0.8466
0.3199	18.99	1187	0.5183	0.8574	0.8552	0.8574	0.8555
0.3121	20.0	1250	0.5367	0.8484	0.8471	0.8484	0.8451
0.2942	20.99	1312	0.5905	0.8534	0.8506	0.8534	0.8509
0.3253	22.0	1375	0.5762	0.8495	0.8498	0.8495	0.8478
0.2917	22.99	1437	0.5865	0.8433	0.8452	0.8433	0.8428
0.2708	24.0	1500	0.5802	0.8568	0.8532	0.8568	0.8539
0.2801	24.99	1562	0.6005	0.8557	0.8521	0.8557	0.8525
0.2608	26.0	1625	0.5916	0.8636	0.8606	0.8636	0.8612
0.2625	26.99	1687	0.5932	0.8568	0.8551	0.8568	0.8552
0.2759	28.0	1750	0.6277	0.8568	0.8557	0.8568	0.8546
0.2483	28.99	1812	0.6055	0.8630	0.8607	0.8630	0.8608
0.2554	29.76	1860	0.6138	0.8653	0.8633	0.8653	0.8633

Framework versions

Transformers 4.37.2
Pytorch 2.3.0
Datasets 2.19.1
Tokenizers 0.15.1