dinov2-base-imagenet1k-1-layer-finetuned-galaxy_mnist

This model is a fine-tuned version of facebook/dinov2-base-imagenet1k-1-layer on the matthieulel/galaxy_mnist dataset. It achieves the following results on the evaluation set:

• Loss: 0.1875
• Accuracy: 0.9365
• Precision: 0.9367
• Recall: 0.9365
• F1: 0.9365

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-06
• 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
0.8401	0.99	31	0.6220	0.7605	0.7579	0.7605	0.7543
0.3857	1.98	62	0.2696	0.8875	0.8881	0.8875	0.8877
0.3144	2.98	93	0.2491	0.9015	0.9028	0.9015	0.9010
0.2769	4.0	125	0.2179	0.913	0.9129	0.913	0.9128
0.2858	4.99	156	0.2455	0.9025	0.9070	0.9025	0.9020
0.2704	5.98	187	0.2121	0.9155	0.9234	0.9155	0.9156
0.2557	6.98	218	0.2177	0.9155	0.9190	0.9155	0.9152
0.2069	8.0	250	0.1864	0.9255	0.9256	0.9255	0.9255
0.2344	8.99	281	0.1894	0.923	0.9237	0.923	0.9230
0.1996	9.98	312	0.1993	0.9235	0.9260	0.9235	0.9234
0.2011	10.98	343	0.1828	0.928	0.9280	0.928	0.9279
0.2229	12.0	375	0.2358	0.9155	0.9233	0.9155	0.9145
0.1792	12.99	406	0.1897	0.9205	0.9214	0.9205	0.9205
0.1898	13.98	437	0.2017	0.921	0.9217	0.921	0.9208
0.1735	14.98	468	0.1954	0.927	0.9270	0.927	0.9269
0.1751	16.0	500	0.1918	0.9295	0.9299	0.9295	0.9294
0.1732	16.99	531	0.1906	0.922	0.9225	0.922	0.9219
0.1738	17.98	562	0.1846	0.931	0.9317	0.931	0.9310
0.1694	18.98	593	0.1875	0.9365	0.9367	0.9365	0.9365
0.1723	20.0	625	0.1941	0.9285	0.9293	0.9285	0.9284
0.1574	20.99	656	0.1905	0.9335	0.9337	0.9335	0.9336
0.1485	21.98	687	0.1869	0.9315	0.9313	0.9315	0.9314
0.1537	22.98	718	0.1830	0.936	0.9360	0.936	0.9360
0.1406	24.0	750	0.1975	0.932	0.9322	0.932	0.9320
0.1326	24.99	781	0.1918	0.9315	0.9316	0.9315	0.9315
0.1238	25.98	812	0.2105	0.9275	0.9288	0.9275	0.9276
0.1299	26.98	843	0.2022	0.9325	0.9327	0.9325	0.9324
0.1387	28.0	875	0.2011	0.9335	0.9337	0.9335	0.9336
0.1279	28.99	906	0.2005	0.931	0.9310	0.931	0.9310
0.1256	29.76	930	0.2004	0.931	0.9310	0.931	0.9310

Framework versions

• Transformers 4.37.2
• Pytorch 2.3.0
• Datasets 2.19.1
• Tokenizers 0.15.1