vit-base-patch16-384-finetuned-humid-binary-2

This model is a fine-tuned version of google/vit-base-patch16-384 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.4611
• Accuracy: 0.9412
• F1 Macro: 0.9399
• Precision Macro: 0.9399
• Recall Macro: 0.9399
• Precision Dry: 0.9310
• Recall Dry: 0.9310
• F1 Dry: 0.9310
• Precision Humid: 0.9487
• Recall Humid: 0.9487
• F1 Humid: 0.9487

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• 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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	Precision Macro	Recall Macro	Precision Dry	Recall Dry	F1 Dry	Precision Humid	Recall Humid	F1 Humid
No log	1.0	6	0.6149	0.7206	0.7176	0.7170	0.7210	0.6562	0.7241	0.6885	0.7778	0.7179	0.7467
0.6101	2.0	12	0.3865	0.8824	0.8798	0.8798	0.8798	0.8621	0.8621	0.8621	0.8974	0.8974	0.8974
0.6101	3.0	18	0.3597	0.8382	0.8379	0.8438	0.8501	0.75	0.9310	0.8308	0.9375	0.7692	0.8451
0.2613	4.0	24	0.5014	0.8235	0.8235	0.8417	0.8417	0.7179	0.9655	0.8235	0.9655	0.7179	0.8235
0.1505	5.0	30	0.4383	0.8529	0.8524	0.8554	0.8630	0.7714	0.9310	0.8438	0.9394	0.7949	0.8611
0.1505	6.0	36	0.4248	0.8529	0.8528	0.8622	0.8674	0.7568	0.9655	0.8485	0.9677	0.7692	0.8571
0.1298	7.0	42	0.3932	0.8971	0.8932	0.9020	0.8882	0.9231	0.8276	0.8727	0.8810	0.9487	0.9136
0.1298	8.0	48	0.3933	0.8971	0.8960	0.8941	0.9014	0.8438	0.9310	0.8852	0.9444	0.8718	0.9067
0.1033	9.0	54	0.7160	0.7941	0.7934	0.8372	0.8205	0.6744	1.0	0.8056	1.0	0.6410	0.7812
0.1135	10.0	60	0.4036	0.9265	0.9252	0.9237	0.9271	0.9	0.9310	0.9153	0.9474	0.9231	0.9351
0.1135	11.0	66	0.4228	0.9118	0.9089	0.9142	0.9054	0.9259	0.8621	0.8929	0.9024	0.9487	0.925
0.0331	12.0	72	0.4821	0.8824	0.8819	0.8848	0.8930	0.8	0.9655	0.875	0.9697	0.8205	0.8889
0.0331	13.0	78	0.4474	0.9118	0.9105	0.9085	0.9142	0.8710	0.9310	0.9	0.9459	0.8974	0.9211
0.0142	14.0	84	0.4478	0.9118	0.9098	0.9098	0.9098	0.8966	0.8966	0.8966	0.9231	0.9231	0.9231
0.0074	15.0	90	0.6302	0.8824	0.8819	0.8848	0.8930	0.8	0.9655	0.875	0.9697	0.8205	0.8889
0.0074	16.0	96	0.4611	0.9412	0.9399	0.9399	0.9399	0.9310	0.9310	0.9310	0.9487	0.9487	0.9487
0.0114	17.0	102	0.6472	0.8824	0.8819	0.8848	0.8930	0.8	0.9655	0.875	0.9697	0.8205	0.8889
0.0114	18.0	108	0.4690	0.9118	0.9089	0.9142	0.9054	0.9259	0.8621	0.8929	0.9024	0.9487	0.925
0.0054	19.0	114	0.7610	0.8676	0.8674	0.8733	0.8802	0.7778	0.9655	0.8615	0.9688	0.7949	0.8732
0.0024	20.0	120	0.5395	0.9118	0.9105	0.9085	0.9142	0.8710	0.9310	0.9	0.9459	0.8974	0.9211
0.0024	21.0	126	0.6275	0.8971	0.8960	0.8941	0.9014	0.8438	0.9310	0.8852	0.9444	0.8718	0.9067
0.0024	22.0	132	0.7972	0.8676	0.8674	0.8733	0.8802	0.7778	0.9655	0.8615	0.9688	0.7949	0.8732
0.0024	23.0	138	0.5466	0.9118	0.9105	0.9085	0.9142	0.8710	0.9310	0.9	0.9459	0.8974	0.9211
0.0009	24.0	144	0.5304	0.8971	0.8960	0.8941	0.9014	0.8438	0.9310	0.8852	0.9444	0.8718	0.9067
0.0012	25.0	150	0.6568	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0012	26.0	156	0.6134	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0004	27.0	162	0.6094	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0004	28.0	168	0.6153	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0003	29.0	174	0.6319	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0002	30.0	180	0.6457	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0002	31.0	186	0.6530	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0002	32.0	192	0.6580	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0002	33.0	198	0.6650	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0002	34.0	204	0.6774	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0002	35.0	210	0.6827	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0002	36.0	216	0.6850	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	37.0	222	0.6846	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	38.0	228	0.6853	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	39.0	234	0.6854	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	40.0	240	0.6845	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	41.0	246	0.6847	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	42.0	252	0.6860	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	43.0	258	0.6875	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	44.0	264	0.6895	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	45.0	270	0.6903	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	46.0	276	0.6914	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	47.0	282	0.6920	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	48.0	288	0.6921	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	49.0	294	0.6923	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041
0.0001	50.0	300	0.6924	0.8971	0.8965	0.8971	0.9058	0.8235	0.9655	0.8889	0.9706	0.8462	0.9041

Framework versions

• Transformers 4.56.1
• Pytorch 2.5.1+cu124
• Datasets 4.0.0
• Tokenizers 0.22.0