swin-base-patch4-window12-384-finetuned-humid-classes-2

This model is a fine-tuned version of microsoft/swin-base-patch4-window12-384 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.1072
• Accuracy: 0.9839
• F1 Macro: 0.9726
• Precision Macro: 0.99
• Recall Macro: 0.96
• Precision Dry: 1.0
• Recall Dry: 1.0
• F1 Dry: 1.0
• Precision Firm: 1.0
• Recall Firm: 1.0
• F1 Firm: 1.0
• Precision Humid: 1.0
• Recall Humid: 0.8
• F1 Humid: 0.8889
• Precision Lump: 0.95
• Recall Lump: 1.0
• F1 Lump: 0.9744
• Precision Rockies: 1.0
• Recall Rockies: 1.0
• F1 Rockies: 1.0

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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	Precision Rockies	Recall Rockies	F1 Rockies
1.4761	1.0	10	1.4416	0.3548	0.2	0.1690	0.2654	0.0	0.0	0.0	0.5625	0.6429	0.6	0.0	0.0	0.0	0.2826	0.6842	0.4	0.0	0.0	0.0
1.0624	2.0	20	0.8707	0.6613	0.5376	0.6525	0.5770	1.0	0.9	0.9474	0.7778	1.0	0.875	0.0	0.0	0.0	0.4848	0.8421	0.6154	1.0	0.1429	0.25
0.4852	3.0	30	0.3910	0.8548	0.7861	0.8978	0.7874	1.0	1.0	1.0	0.7778	1.0	0.875	1.0	0.2	0.3333	0.7778	0.7368	0.7568	0.9333	1.0	0.9655
0.3417	4.0	40	0.4915	0.8548	0.7875	0.9167	0.7761	1.0	1.0	1.0	0.875	1.0	0.9333	1.0	0.2	0.3333	0.7083	0.8947	0.7907	1.0	0.7857	0.88
0.175	5.0	50	0.1700	0.9355	0.9498	0.95	0.9579	1.0	1.0	1.0	0.875	1.0	0.9333	1.0	1.0	1.0	1.0	0.7895	0.8824	0.875	1.0	0.9333
0.161	6.0	60	0.1708	0.9194	0.9242	0.9431	0.9179	1.0	1.0	1.0	0.7778	1.0	0.875	1.0	0.8	0.8889	0.9375	0.7895	0.8571	1.0	1.0	1.0
0.2785	7.0	70	0.1375	0.9355	0.9381	0.9283	0.9579	1.0	1.0	1.0	0.875	1.0	0.9333	0.8333	1.0	0.9091	1.0	0.7895	0.8824	0.9333	1.0	0.9655
0.1098	8.0	80	0.1949	0.9194	0.9026	0.8980	0.9179	1.0	1.0	1.0	1.0	1.0	1.0	0.6667	0.8	0.7273	1.0	0.7895	0.8824	0.8235	1.0	0.9032
0.1128	9.0	90	0.0844	0.9355	0.9147	0.9133	0.9209	1.0	1.0	1.0	1.0	1.0	1.0	0.6667	0.8	0.7273	0.9	0.9474	0.9231	1.0	0.8571	0.9231
0.0714	10.0	100	0.1127	0.9677	0.9604	0.9810	0.9457	1.0	1.0	1.0	1.0	1.0	1.0	1.0	0.8	0.8889	0.9048	1.0	0.95	1.0	0.9286	0.9630
0.0585	11.0	110	0.2509	0.9677	0.9400	0.9810	0.9200	1.0	1.0	1.0	1.0	1.0	1.0	1.0	0.6	0.75	0.9048	1.0	0.95	1.0	1.0	1.0
0.0565	12.0	120	0.0925	0.9677	0.9613	0.9567	0.9714	1.0	1.0	1.0	1.0	1.0	1.0	0.8333	1.0	0.9091	0.95	1.0	0.9744	1.0	0.8571	0.9231
0.0192	13.0	130	0.1702	0.9677	0.9552	0.9767	0.9400	1.0	0.9	0.9474	1.0	1.0	1.0	1.0	0.8	0.8889	0.95	1.0	0.9744	0.9333	1.0	0.9655
0.0738	14.0	140	0.3350	0.9194	0.9008	0.9	0.9398	1.0	1.0	1.0	1.0	1.0	1.0	0.5	1.0	0.6667	1.0	0.8421	0.9143	1.0	0.8571	0.9231
0.0447	15.0	150	0.2240	0.9355	0.9183	0.9111	0.9541	1.0	1.0	1.0	1.0	1.0	1.0	0.5556	1.0	0.7143	1.0	0.8421	0.9143	1.0	0.9286	0.9630
0.0189	16.0	160	0.1072	0.9839	0.9726	0.99	0.96	1.0	1.0	1.0	1.0	1.0	1.0	1.0	0.8	0.8889	0.95	1.0	0.9744	1.0	1.0	1.0
0.0112	17.0	170	0.9942	0.8871	0.8116	0.9462	0.7971	1.0	1.0	1.0	1.0	1.0	1.0	1.0	0.2	0.3333	0.7308	1.0	0.8444	1.0	0.7857	0.88
0.0921	18.0	180	0.2717	0.9516	0.9478	0.9727	0.9314	1.0	1.0	1.0	1.0	1.0	1.0	1.0	0.8	0.8889	0.8636	1.0	0.9268	1.0	0.8571	0.9231
0.0042	19.0	190	0.0812	0.9677	0.9495	0.9495	0.9495	1.0	1.0	1.0	1.0	1.0	1.0	0.8	0.8	0.8	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0016	20.0	200	0.1383	0.9839	0.9726	0.99	0.96	1.0	1.0	1.0	1.0	1.0	1.0	1.0	0.8	0.8889	0.95	1.0	0.9744	1.0	1.0	1.0
0.0103	21.0	210	0.2240	0.9516	0.9277	0.9667	0.9095	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.6	0.75	0.9	0.9474	0.9231	1.0	1.0	1.0
0.0	22.0	220	0.1629	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	23.0	230	0.1367	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0001	24.0	240	0.1385	0.9516	0.9352	0.9361	0.9352	1.0	1.0	1.0	0.9333	1.0	0.9655	0.8	0.8	0.8	0.9474	0.9474	0.9474	1.0	0.9286	0.9630
0.0003	25.0	250	0.1683	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0016	26.0	260	0.1597	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0001	27.0	270	0.2348	0.9516	0.9346	0.9410	0.9314	1.0	1.0	1.0	1.0	1.0	1.0	0.8	0.8	0.8	0.9048	1.0	0.95	1.0	0.8571	0.9231
0.0001	28.0	280	0.0697	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	29.0	290	0.0645	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0003	30.0	300	0.3096	0.9516	0.9277	0.9667	0.9095	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.6	0.75	0.9	0.9474	0.9231	1.0	1.0	1.0
0.0003	31.0	310	0.2795	0.9516	0.9277	0.9667	0.9095	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.6	0.75	0.9	0.9474	0.9231	1.0	1.0	1.0
0.0	32.0	320	0.1708	0.9677	0.9638	0.9533	0.9789	1.0	1.0	1.0	0.9333	1.0	0.9655	0.8333	1.0	0.9091	1.0	0.8947	0.9444	1.0	1.0	1.0
0.0002	33.0	330	0.1291	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	34.0	340	0.2383	0.9516	0.9277	0.9667	0.9095	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.6	0.75	0.9	0.9474	0.9231	1.0	1.0	1.0
0.0	35.0	350	0.2886	0.9516	0.9277	0.9667	0.9095	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.6	0.75	0.9	0.9474	0.9231	1.0	1.0	1.0
0.0006	36.0	360	0.0748	0.9839	0.9877	0.9867	0.9895	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	1.0	1.0	1.0	0.9474	0.9730	1.0	1.0	1.0
0.0001	37.0	370	0.1293	0.9677	0.9638	0.9533	0.9789	1.0	1.0	1.0	0.9333	1.0	0.9655	0.8333	1.0	0.9091	1.0	0.8947	0.9444	1.0	1.0	1.0
0.0	38.0	380	0.0945	0.9677	0.9638	0.9533	0.9789	1.0	1.0	1.0	0.9333	1.0	0.9655	0.8333	1.0	0.9091	1.0	0.8947	0.9444	1.0	1.0	1.0
0.0	39.0	390	0.0562	0.9677	0.9638	0.9533	0.9789	1.0	1.0	1.0	0.9333	1.0	0.9655	0.8333	1.0	0.9091	1.0	0.8947	0.9444	1.0	1.0	1.0
0.0	40.0	400	0.0382	0.9839	0.9877	0.9867	0.9895	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	1.0	1.0	1.0	0.9474	0.9730	1.0	1.0	1.0
0.0	41.0	410	0.0368	0.9839	0.9877	0.9867	0.9895	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	1.0	1.0	1.0	0.9474	0.9730	1.0	1.0	1.0
0.0	42.0	420	0.0566	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	43.0	430	0.0815	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	44.0	440	0.0899	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	45.0	450	0.0921	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	46.0	460	0.0914	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	47.0	470	0.0896	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0001	48.0	480	0.0890	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0	49.0	490	0.0903	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0
0.0001	50.0	500	0.0914	0.9677	0.9604	0.9761	0.9495	1.0	1.0	1.0	0.9333	1.0	0.9655	1.0	0.8	0.8889	0.9474	0.9474	0.9474	1.0	1.0	1.0

Framework versions

• Transformers 4.56.1
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.0