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: 0.3769
Accuracy: 0.9241
F1 Macro: 0.8482
Precision Macro: 0.9434
Recall Macro: 0.8204
Precision Dry: 0.8889
Recall Dry: 1.0
F1 Dry: 0.9412
Precision Firm: 1.0
Recall Firm: 0.9231
F1 Firm: 0.96
Precision Humid: 1.0
Recall Humid: 0.4
F1 Humid: 0.5714
Precision Lump: 0.8846
Recall Lump: 0.9583
F1 Lump: 0.92

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 Firm	Recall Firm	F1 Firm	Precision Humid	Recall Humid	F1 Humid	Precision Lump	Recall Lump	F1 Lump
No log	1.0	7	1.1773	0.5190	0.3767	0.4228	0.4087	0.6667	0.25	0.3636	0.4792	0.8846	0.6216	0.0	0.0	0.0	0.5455	0.5	0.5217
1.2739	2.0	14	0.6848	0.8608	0.6673	0.6497	0.6891	0.88	0.9167	0.8980	0.96	0.9231	0.9412	0.0	0.0	0.0	0.7586	0.9167	0.8302
0.6466	3.0	21	0.4529	0.8481	0.6590	0.6497	0.6803	0.88	0.9167	0.8980	1.0	0.8462	0.9167	0.0	0.0	0.0	0.7188	0.9583	0.8214
0.6466	4.0	28	0.3529	0.8861	0.6861	0.6658	0.7099	0.8571	1.0	0.9231	0.96	0.9231	0.9412	0.0	0.0	0.0	0.8462	0.9167	0.88
0.3484	5.0	35	0.3661	0.8608	0.6671	0.6560	0.6915	0.8571	1.0	0.9231	1.0	0.8077	0.8936	0.0	0.0	0.0	0.7667	0.9583	0.8519
0.2104	6.0	42	0.2790	0.8987	0.7731	0.9223	0.7583	0.88	0.9167	0.8980	0.9630	1.0	0.9811	1.0	0.2	0.3333	0.8462	0.9167	0.88
0.2104	7.0	49	0.2667	0.9114	0.7832	0.9325	0.7688	0.8846	0.9583	0.92	0.9286	1.0	0.9630	1.0	0.2	0.3333	0.9167	0.9167	0.9167
0.1791	8.0	56	0.3391	0.8608	0.7944	0.8058	0.8063	0.8846	0.9583	0.92	0.8387	1.0	0.9123	0.5	0.6	0.5455	1.0	0.6667	0.8
0.0883	9.0	63	0.3446	0.8861	0.8014	0.8419	0.7875	0.8571	1.0	0.9231	0.8966	1.0	0.9455	0.6667	0.4	0.5	0.9474	0.75	0.8372
0.0567	10.0	70	0.2870	0.8987	0.8280	0.8251	0.8383	0.8889	1.0	0.9412	0.9615	0.9615	0.9615	0.5	0.6	0.5455	0.95	0.7917	0.8636
0.0567	11.0	77	0.3291	0.8987	0.8287	0.9265	0.8003	0.8846	0.9583	0.92	1.0	0.8846	0.9388	1.0	0.4	0.5714	0.8214	0.9583	0.8846
0.0475	12.0	84	0.2933	0.8987	0.8519	0.8716	0.8399	0.9565	0.9167	0.9362	0.9583	0.8846	0.92	0.75	0.6	0.6667	0.8214	0.9583	0.8846
0.0501	13.0	91	0.3152	0.8861	0.8168	0.8155	0.8279	0.8889	1.0	0.9412	0.9259	0.9615	0.9434	0.5	0.6	0.5455	0.9474	0.75	0.8372
0.0501	14.0	98	0.2582	0.8987	0.8282	0.9227	0.7995	0.8846	0.9583	0.92	0.96	0.9231	0.9412	1.0	0.4	0.5714	0.8462	0.9167	0.88
0.0267	15.0	105	0.2268	0.9114	0.8605	0.8773	0.8487	0.8846	0.9583	0.92	0.9615	0.9615	0.9615	0.75	0.6	0.6667	0.9130	0.875	0.8936
0.0129	16.0	112	0.3056	0.9114	0.8753	0.8662	0.8899	0.9231	1.0	0.96	1.0	0.8846	0.9388	0.6667	0.8	0.7273	0.875	0.875	0.875
0.0129	17.0	119	0.3527	0.8987	0.8803	0.8856	0.8795	0.9565	0.9167	0.9362	1.0	0.8846	0.9388	0.8	0.8	0.8	0.7857	0.9167	0.8462
0.0168	18.0	126	0.3196	0.8987	0.8280	0.8251	0.8383	0.8889	1.0	0.9412	0.9615	0.9615	0.9615	0.5	0.6	0.5455	0.95	0.7917	0.8636
0.0145	19.0	133	0.4022	0.8861	0.7954	0.8072	0.7899	0.8889	1.0	0.9412	1.0	0.8846	0.9388	0.5	0.4	0.4444	0.84	0.875	0.8571
0.0078	20.0	140	0.3966	0.8861	0.7934	0.8059	0.7883	0.8571	1.0	0.9231	0.9615	0.9615	0.9615	0.5	0.4	0.4444	0.9048	0.7917	0.8444
0.0078	21.0	147	0.3966	0.8734	0.8096	0.8058	0.8199	0.9231	1.0	0.96	1.0	0.8462	0.9167	0.5	0.6	0.5455	0.8	0.8333	0.8163
0.0056	22.0	154	0.3384	0.8987	0.8143	0.8476	0.7987	0.8846	0.9583	0.92	0.9259	0.9615	0.9434	0.6667	0.4	0.5	0.9130	0.875	0.8936
0.0043	23.0	161	0.3612	0.9114	0.8379	0.9322	0.8099	0.8889	1.0	0.9412	0.96	0.9231	0.9412	1.0	0.4	0.5714	0.88	0.9167	0.8980
0.0043	24.0	168	0.4534	0.9114	0.8383	0.9352	0.8107	0.8889	1.0	0.9412	1.0	0.8846	0.9388	1.0	0.4	0.5714	0.8519	0.9583	0.9020
0.0044	25.0	175	0.3895	0.8987	0.8052	0.8160	0.7995	0.8889	1.0	0.9412	1.0	0.9231	0.96	0.5	0.4	0.4444	0.875	0.875	0.875
0.0051	26.0	182	0.4065	0.8987	0.8042	0.8149	0.7987	0.8889	1.0	0.9412	0.9615	0.9615	0.9615	0.5	0.4	0.4444	0.9091	0.8333	0.8696
0.0051	27.0	189	0.5653	0.8861	0.8416	0.8647	0.8311	0.9231	1.0	0.96	1.0	0.8077	0.8936	0.75	0.6	0.6667	0.7857	0.9167	0.8462
0.004	28.0	196	0.3769	0.9241	0.8482	0.9434	0.8204	0.8889	1.0	0.9412	1.0	0.9231	0.96	1.0	0.4	0.5714	0.8846	0.9583	0.92
0.0065	29.0	203	0.3687	0.9241	0.8338	0.8683	0.8187	0.8889	1.0	0.9412	0.9630	1.0	0.9811	0.6667	0.4	0.5	0.9545	0.875	0.9130
0.0038	30.0	210	0.4031	0.9114	0.8488	0.8495	0.8495	0.9231	1.0	0.96	1.0	0.9231	0.96	0.6	0.6	0.6	0.875	0.875	0.875
0.0038	31.0	217	0.5323	0.8987	0.8390	0.8408	0.8399	0.9231	1.0	0.96	1.0	0.8846	0.9388	0.6	0.6	0.6	0.84	0.875	0.8571
0.0136	32.0	224	0.4092	0.8987	0.8042	0.8149	0.7987	0.8889	1.0	0.9412	0.9615	0.9615	0.9615	0.5	0.4	0.4444	0.9091	0.8333	0.8696
0.0054	33.0	231	0.3930	0.9114	0.8248	0.8589	0.8099	0.8889	1.0	0.9412	1.0	0.9231	0.96	0.6667	0.4	0.5	0.88	0.9167	0.8980
0.0054	34.0	238	0.5258	0.8987	0.8150	0.8504	0.8003	0.8889	1.0	0.9412	1.0	0.8846	0.9388	0.6667	0.4	0.5	0.8462	0.9167	0.88
0.0038	35.0	245	0.4985	0.8987	0.8150	0.8504	0.8003	0.8889	1.0	0.9412	1.0	0.8846	0.9388	0.6667	0.4	0.5	0.8462	0.9167	0.88
0.0059	36.0	252	0.4084	0.9114	0.8241	0.8575	0.8091	0.8889	1.0	0.9412	0.9615	0.9615	0.9615	0.6667	0.4	0.5	0.9130	0.875	0.8936
0.0059	37.0	259	0.3682	0.9241	0.8346	0.8681	0.8196	0.8889	1.0	0.9412	1.0	0.9615	0.9804	0.6667	0.4	0.5	0.9167	0.9167	0.9167
0.004	38.0	266	0.3731	0.9114	0.8248	0.8589	0.8099	0.8889	1.0	0.9412	1.0	0.9231	0.96	0.6667	0.4	0.5	0.88	0.9167	0.8980
0.0028	39.0	273	0.3802	0.9241	0.8712	0.8883	0.8599	0.9231	1.0	0.96	1.0	0.9231	0.96	0.75	0.6	0.6667	0.88	0.9167	0.8980
0.004	40.0	280	0.3915	0.9114	0.8488	0.8495	0.8495	0.9231	1.0	0.96	1.0	0.9231	0.96	0.6	0.6	0.6	0.875	0.875	0.875
0.004	41.0	287	0.3963	0.9114	0.8488	0.8495	0.8495	0.9231	1.0	0.96	1.0	0.9231	0.96	0.6	0.6	0.6	0.875	0.875	0.875
0.0024	42.0	294	0.3943	0.9114	0.8488	0.8495	0.8495	0.9231	1.0	0.96	1.0	0.9231	0.96	0.6	0.6	0.6	0.875	0.875	0.875
0.0053	43.0	301	0.3857	0.9114	0.8488	0.8495	0.8495	0.9231	1.0	0.96	1.0	0.9231	0.96	0.6	0.6	0.6	0.875	0.875	0.875
0.0053	44.0	308	0.3823	0.9241	0.8712	0.8883	0.8599	0.9231	1.0	0.96	1.0	0.9231	0.96	0.75	0.6	0.6667	0.88	0.9167	0.8980
0.0042	45.0	315	0.3791	0.9114	0.8248	0.8589	0.8099	0.8889	1.0	0.9412	1.0	0.9231	0.96	0.6667	0.4	0.5	0.88	0.9167	0.8980
0.0037	46.0	322	0.3782	0.9114	0.8248	0.8589	0.8099	0.8889	1.0	0.9412	1.0	0.9231	0.96	0.6667	0.4	0.5	0.88	0.9167	0.8980
0.0037	47.0	329	0.3785	0.9241	0.8712	0.8883	0.8599	0.9231	1.0	0.96	1.0	0.9231	0.96	0.75	0.6	0.6667	0.88	0.9167	0.8980
0.0026	48.0	336	0.3769	0.9114	0.8248	0.8589	0.8099	0.8889	1.0	0.9412	1.0	0.9231	0.96	0.6667	0.4	0.5	0.88	0.9167	0.8980
0.0051	49.0	343	0.3763	0.9114	0.8248	0.8589	0.8099	0.8889	1.0	0.9412	1.0	0.9231	0.96	0.6667	0.4	0.5	0.88	0.9167	0.8980
0.0025	50.0	350	0.3763	0.9114	0.8248	0.8589	0.8099	0.8889	1.0	0.9412	1.0	0.9231	0.96	0.6667	0.4	0.5	0.88	0.9167	0.8980

Framework versions

Transformers 4.56.1
Pytorch 2.5.1+cu124
Datasets 4.0.0
Tokenizers 0.22.0

dacunaq/vit-base-patch16-384-finetuned-humid-classes-1

vit-base-patch16-384-finetuned-humid-classes-1

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

dacunaq/
vit-base-patch16-384-finetuned-humid-classes-1