Dinotron

This model is a fine-tuned version of facebook/dinov2-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0265
• Accuracy: 0.9932

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: 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	7	0.1146	0.9638
0.3773	2.0	14	0.0336	0.9932
0.0541	3.0	21	0.0402	0.9887
0.0541	4.0	28	0.0463	0.9887
0.0476	5.0	35	0.0594	0.9819
0.1408	6.0	42	0.1296	0.9570
0.1408	7.0	49	0.0872	0.9729
0.0898	8.0	56	0.2245	0.9344
0.216	9.0	63	0.1444	0.9570
0.076	10.0	70	0.0316	0.9887
0.076	11.0	77	0.0411	0.9864
0.0369	12.0	84	0.0275	0.9887
0.0505	13.0	91	0.1610	0.9638
0.0505	14.0	98	0.0513	0.9910
0.0274	15.0	105	0.2366	0.9615
0.0735	16.0	112	0.0738	0.9796
0.0735	17.0	119	0.0529	0.9819
0.0334	18.0	126	0.1024	0.9661
0.0347	19.0	133	0.0919	0.9819
0.0206	20.0	140	0.0851	0.9864
0.0206	21.0	147	0.1004	0.9796
0.0516	22.0	154	0.1706	0.9638
0.0418	23.0	161	0.0505	0.9910
0.0418	24.0	168	0.0939	0.9774
0.0173	25.0	175	0.0553	0.9842
0.0239	26.0	182	0.1255	0.9796
0.0239	27.0	189	0.2256	0.9661
0.0286	28.0	196	0.0943	0.9751
0.0502	29.0	203	0.0937	0.9751
0.0102	30.0	210	0.0910	0.9842
0.0102	31.0	217	0.0336	0.9887
0.0182	32.0	224	0.0870	0.9796
0.0126	33.0	231	0.0565	0.9842
0.0126	34.0	238	0.0541	0.9842
0.0157	35.0	245	0.0591	0.9932
0.0059	36.0	252	0.0985	0.9819
0.0059	37.0	259	0.0813	0.9819
0.0092	38.0	266	0.0239	0.9955
0.0225	39.0	273	0.0982	0.9706
0.0105	40.0	280	0.0113	0.9955
0.0105	41.0	287	0.0127	0.9977
0.007	42.0	294	0.0760	0.9887
0.0032	43.0	301	0.0196	0.9932
0.0032	44.0	308	0.0171	0.9932
0.0206	45.0	315	0.0501	0.9910
0.0001	46.0	322	0.0925	0.9842
0.0001	47.0	329	0.0318	0.9910
0.0017	48.0	336	0.0612	0.9864
0.0023	49.0	343	0.0685	0.9864
0.0013	50.0	350	0.0265	0.9932

Framework versions

• Transformers 4.33.3
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3