dinov2-base-imagenet1k-1-layer-finetuned-galaxy10-decals-head-finetuned-30-galaxy_mnist

This model is a fine-tuned version of matthieulel/dinov2-base-imagenet1k-1-layer-finetuned-galaxy10-decals on the matthieulel/galaxy_mnist dataset. It achieves the following results on the evaluation set:

• Loss: 0.2139
• Accuracy: 0.9085
• Precision: 0.9092
• Recall: 0.9085
• F1: 0.9085

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
1.1345	0.99	62	0.9672	0.602	0.6135	0.602	0.5997
0.5365	2.0	125	0.3855	0.8585	0.8602	0.8585	0.8585
0.3715	2.99	187	0.2806	0.8875	0.8873	0.8875	0.8872
0.3509	4.0	250	0.2557	0.895	0.8949	0.895	0.8949
0.3333	4.99	312	0.2504	0.898	0.8985	0.898	0.8979
0.3073	6.0	375	0.2424	0.8965	0.8979	0.8965	0.8964
0.3169	6.99	437	0.2352	0.898	0.8982	0.898	0.8979
0.2698	8.0	500	0.2348	0.896	0.8971	0.896	0.8959
0.2681	8.99	562	0.2300	0.903	0.9040	0.903	0.9030
0.2533	10.0	625	0.2274	0.9045	0.9049	0.9045	0.9045
0.2816	10.99	687	0.2247	0.9035	0.9037	0.9035	0.9034
0.27	12.0	750	0.2227	0.9055	0.9057	0.9055	0.9055
0.2716	12.99	812	0.2240	0.903	0.9035	0.903	0.9030
0.2638	14.0	875	0.2200	0.904	0.9042	0.904	0.9040
0.3064	14.99	937	0.2222	0.907	0.9083	0.907	0.9069
0.268	16.0	1000	0.2215	0.9065	0.9083	0.9065	0.9064
0.2717	16.99	1062	0.2192	0.907	0.9080	0.907	0.9070
0.2706	18.0	1125	0.2185	0.907	0.9080	0.907	0.9070
0.2697	18.99	1187	0.2184	0.907	0.9084	0.907	0.9070
0.2696	20.0	1250	0.2169	0.906	0.9071	0.906	0.9060
0.2966	20.99	1312	0.2158	0.9055	0.9064	0.9055	0.9055
0.273	22.0	1375	0.2175	0.9065	0.9079	0.9065	0.9064
0.259	22.99	1437	0.2183	0.904	0.9059	0.904	0.9039
0.2711	24.0	1500	0.2167	0.906	0.9074	0.906	0.9060
0.2704	24.99	1562	0.2160	0.9065	0.9081	0.9065	0.9064
0.2233	26.0	1625	0.2139	0.9085	0.9092	0.9085	0.9085
0.2759	26.99	1687	0.2152	0.907	0.9084	0.907	0.9070
0.2667	28.0	1750	0.2156	0.907	0.9085	0.907	0.9070
0.2663	28.99	1812	0.2156	0.907	0.9085	0.907	0.9070
0.2645	29.76	1860	0.2154	0.907	0.9085	0.907	0.9070

Framework versions

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