DinoAmoros-small-2025_05_06_34659-prova_bs16_freeze_monolabel

This model is a fine-tuned version of facebook/dinov2-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.1588
• F1 Micro: 0.2
• F1 Macro: 0.1339
• Accuracy: 0.2
• Learning Rate: 0.0001

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: 0.001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 150
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Micro	F1 Macro	Accuracy	Rate
No log	1.0	1	3.3289	0.0	0.0	0.0	0.001
No log	2.0	2	3.2965	0.0	0.0	0.0	0.001
No log	3.0	3	3.2832	0.0	0.0	0.0	0.001
No log	4.0	4	3.2658	0.0	0.0	0.0	0.001
No log	5.0	5	3.2574	0.0	0.0	0.0	0.001
No log	6.0	6	3.2430	0.1	0.0444	0.1	0.001
No log	7.0	7	3.2252	0.0	0.0	0.0	0.001
No log	8.0	8	3.2078	0.0	0.0	0.0	0.001
No log	9.0	9	3.2320	0.0	0.0	0.0	0.001
No log	10.0	10	3.2318	0.0	0.0	0.0	0.001
No log	11.0	11	3.2295	0.1	0.0417	0.1	0.001
No log	12.0	12	3.1906	0.3	0.3254	0.3	0.001
No log	13.0	13	3.1611	0.5	0.4278	0.5	0.001
No log	14.0	14	3.1342	0.5	0.4278	0.5	0.001
No log	15.0	15	3.1030	0.4	0.3333	0.4	0.001
No log	16.0	16	3.0616	0.4	0.3333	0.4	0.001
No log	17.0	17	3.0400	0.4	0.3333	0.4	0.001
No log	18.0	18	2.9770	0.4	0.3810	0.4	0.001
No log	19.0	19	2.9270	0.4	0.3810	0.4	0.001
No log	20.0	20	2.8651	0.5	0.5	0.5	0.001
No log	21.0	21	2.8102	0.5	0.5	0.5	0.001
No log	22.0	22	2.7777	0.5	0.6111	0.5	0.001
No log	23.0	23	2.7405	0.5	0.5	0.5	0.001
No log	24.0	24	2.6667	0.5	0.6111	0.5	0.001
No log	25.0	25	2.6132	0.5	0.6111	0.5	0.001
No log	26.0	26	2.5834	0.5	0.5	0.5	0.001
No log	27.0	27	2.5603	0.5	0.5	0.5	0.001
No log	28.0	28	2.5502	0.5	0.5	0.5	0.001
No log	29.0	29	2.5178	0.5	0.5	0.5	0.001
No log	30.0	30	2.4880	0.5	0.5	0.5	0.001
No log	31.0	31	2.4448	0.5	0.6111	0.5	0.001
No log	32.0	32	2.4291	0.5	0.6111	0.5	0.001
No log	33.0	33	2.4009	0.5	0.6111	0.5	0.001
No log	34.0	34	2.3793	0.5	0.6111	0.5	0.001
No log	35.0	35	2.3791	0.5	0.6111	0.5	0.001
No log	36.0	36	2.3485	0.5	0.6111	0.5	0.001
No log	37.0	37	2.3341	0.5	0.6111	0.5	0.001
No log	38.0	38	2.3232	0.5	0.6111	0.5	0.001
No log	39.0	39	2.3187	0.5	0.6111	0.5	0.001
No log	40.0	40	2.3267	0.5	0.6111	0.5	0.001
No log	41.0	41	2.3401	0.5	0.6111	0.5	0.001
No log	42.0	42	2.3547	0.5	0.6111	0.5	0.001
No log	43.0	43	2.3599	0.5	0.6111	0.5	0.001
No log	44.0	44	2.3857	0.5	0.6111	0.5	0.001
No log	45.0	45	2.4031	0.5	0.6111	0.5	0.001
No log	46.0	46	2.3779	0.5	0.6111	0.5	0.0001
No log	47.0	47	2.3591	0.5	0.6111	0.5	0.0001
No log	48.0	48	2.3669	0.5	0.6111	0.5	0.0001
No log	49.0	49	2.3603	0.5	0.6111	0.5	0.0001

Framework versions

• Transformers 4.48.0
• Pytorch 2.6.0+cu118
• Datasets 3.0.2
• Tokenizers 0.21.1