metadata

license: apache-2.0
base_model: facebook/convnextv2-nano-22k-384
tags:
  - image-classification
  - vision
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: convnextv2-nano-22k-384-finetuned-galaxy10-decals
    results: []

convnextv2-nano-22k-384-finetuned-galaxy10-decals

This model is a fine-tuned version of facebook/convnextv2-nano-22k-384 on the matthieulel/galaxy10_decals dataset. It achieves the following results on the evaluation set:

Loss: 0.4057
Accuracy: 0.8681
Precision: 0.8662
Recall: 0.8681
F1: 0.8650

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
1.6939	0.99	62	1.5326	0.4656	0.4580	0.4656	0.4176
0.9882	2.0	125	0.8491	0.7142	0.7196	0.7142	0.7066
0.7595	2.99	187	0.6041	0.7993	0.7990	0.7993	0.7947
0.6097	4.0	250	0.5397	0.8134	0.8078	0.8134	0.8069
0.5565	4.99	312	0.4990	0.8286	0.8269	0.8286	0.8268
0.5822	6.0	375	0.4684	0.8427	0.8425	0.8427	0.8374
0.5244	6.99	437	0.4484	0.8512	0.8476	0.8512	0.8483
0.4957	8.0	500	0.4487	0.8506	0.8543	0.8506	0.8514
0.4857	8.99	562	0.4369	0.8579	0.8572	0.8579	0.8545
0.4634	10.0	625	0.4104	0.8658	0.8630	0.8658	0.8639
0.4433	10.99	687	0.4117	0.8664	0.8649	0.8664	0.8651
0.4267	12.0	750	0.4096	0.8664	0.8632	0.8664	0.8634
0.4201	12.99	812	0.4212	0.8658	0.8645	0.8658	0.8631
0.4176	14.0	875	0.4057	0.8681	0.8662	0.8681	0.8650
0.3717	14.99	937	0.4299	0.8568	0.8547	0.8568	0.8551
0.3759	16.0	1000	0.4446	0.8585	0.8563	0.8585	0.8555
0.3264	16.99	1062	0.4276	0.8647	0.8630	0.8647	0.8623
0.3573	18.0	1125	0.4199	0.8641	0.8621	0.8641	0.8610
0.3356	18.99	1187	0.4388	0.8585	0.8597	0.8585	0.8579
0.3313	20.0	1250	0.4385	0.8602	0.8585	0.8602	0.8571
0.3044	20.99	1312	0.4485	0.8585	0.8578	0.8585	0.8560
0.3525	22.0	1375	0.4303	0.8647	0.8641	0.8647	0.8634
0.3207	22.99	1437	0.4525	0.8608	0.8597	0.8608	0.8591
0.3044	24.0	1500	0.4417	0.8591	0.8578	0.8591	0.8579
0.3088	24.99	1562	0.4626	0.8608	0.8586	0.8608	0.8582
0.2897	26.0	1625	0.4524	0.8630	0.8606	0.8630	0.8606
0.2823	26.99	1687	0.4433	0.8670	0.8657	0.8670	0.8657
0.2928	28.0	1750	0.4479	0.8658	0.8629	0.8658	0.8631
0.2695	28.99	1812	0.4455	0.8658	0.8637	0.8658	0.8639
0.274	29.76	1860	0.4449	0.8630	0.8607	0.8630	0.8610

Framework versions

Transformers 4.37.2
Pytorch 2.3.0
Datasets 2.19.1
Tokenizers 0.15.1