swin-tiny-patch4-window7-224-finetuned-woody_90epochs

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.4351
Accuracy: 0.8424

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: 90

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6659	1.0	58	0.6216	0.6558
0.6181	2.0	116	0.5616	0.7115
0.5941	3.0	174	0.5464	0.7224
0.5727	4.0	232	0.5368	0.7297
0.573	5.0	290	0.4971	0.7539
0.5724	6.0	348	0.4920	0.7467
0.5584	7.0	406	0.4949	0.7564
0.5352	8.0	464	0.5255	0.7406
0.5857	9.0	522	0.4954	0.7515
0.5352	10.0	580	0.4888	0.7455
0.5161	11.0	638	0.5306	0.7224
0.5457	12.0	696	0.4856	0.76
0.5309	13.0	754	0.4647	0.7612
0.5357	14.0	812	0.4688	0.7697
0.5183	15.0	870	0.4830	0.7527
0.4837	16.0	928	0.5238	0.7370
0.51	17.0	986	0.4658	0.7745
0.533	18.0	1044	0.4589	0.7673
0.4808	19.0	1102	0.4375	0.7794
0.4854	20.0	1160	0.4574	0.7745
0.4708	21.0	1218	0.4738	0.7709
0.4801	22.0	1276	0.4688	0.76
0.4751	23.0	1334	0.4610	0.7648
0.497	24.0	1392	0.5058	0.7624
0.4767	25.0	1450	0.4709	0.7721
0.4805	26.0	1508	0.4447	0.7697
0.4557	27.0	1566	0.4558	0.7721
0.4636	28.0	1624	0.4325	0.8036
0.4285	29.0	1682	0.4526	0.7794
0.4358	30.0	1740	0.4302	0.8048
0.4257	31.0	1798	0.4373	0.7927
0.4137	32.0	1856	0.4458	0.7903
0.4389	33.0	1914	0.4522	0.7988
0.4537	34.0	1972	0.4395	0.7927
0.4249	35.0	2030	0.4348	0.8
0.4244	36.0	2088	0.4650	0.7867
0.4256	37.0	2146	0.4402	0.8012
0.4118	38.0	2204	0.4394	0.7867
0.4128	39.0	2262	0.4225	0.8133
0.416	40.0	2320	0.4410	0.8073
0.4211	41.0	2378	0.4464	0.8024
0.3838	42.0	2436	0.4440	0.7976
0.374	43.0	2494	0.4175	0.7903
0.412	44.0	2552	0.4169	0.8109
0.3746	45.0	2610	0.4243	0.8012
0.3719	46.0	2668	0.4132	0.8242
0.381	47.0	2726	0.4485	0.7988
0.3708	48.0	2784	0.4200	0.8085
0.3591	49.0	2842	0.4071	0.8279
0.3762	50.0	2900	0.4428	0.8145
0.3426	51.0	2958	0.4058	0.8158
0.3541	52.0	3016	0.4470	0.8182
0.3373	53.0	3074	0.4252	0.8194
0.3303	54.0	3132	0.4040	0.8315
0.3275	55.0	3190	0.4235	0.8291
0.3151	56.0	3248	0.3984	0.8485
0.324	57.0	3306	0.4283	0.8291
0.3276	58.0	3364	0.4731	0.8145
0.3208	59.0	3422	0.4360	0.8255
0.3355	60.0	3480	0.4143	0.8230
0.3154	61.0	3538	0.4234	0.8267
0.3451	62.0	3596	0.4059	0.8242
0.3071	63.0	3654	0.3991	0.8267
0.3303	64.0	3712	0.4099	0.8242
0.29	65.0	3770	0.4140	0.8327
0.2937	66.0	3828	0.4590	0.8218
0.3322	67.0	3886	0.4111	0.8327
0.3219	68.0	3944	0.4299	0.8327
0.2839	69.0	4002	0.4074	0.8424
0.2903	70.0	4060	0.4366	0.8315
0.2851	71.0	4118	0.4132	0.8473
0.3029	72.0	4176	0.4239	0.8473
0.2693	73.0	4234	0.4194	0.8412
0.2715	74.0	4292	0.4384	0.8412
0.2842	75.0	4350	0.4279	0.8448
0.2733	76.0	4408	0.4174	0.84
0.2694	77.0	4466	0.3966	0.8388
0.2527	78.0	4524	0.4194	0.8364
0.2813	79.0	4582	0.4231	0.8436
0.2618	80.0	4640	0.4494	0.8352
0.2639	81.0	4698	0.4152	0.8388
0.2643	82.0	4756	0.4241	0.8448
0.276	83.0	4814	0.4518	0.8327
0.2761	84.0	4872	0.4349	0.8412
0.2295	85.0	4930	0.4504	0.8315
0.2723	86.0	4988	0.4385	0.8388
0.2559	87.0	5046	0.4362	0.8473
0.2583	88.0	5104	0.4273	0.8436
0.2523	89.0	5162	0.4292	0.8424
0.2563	90.0	5220	0.4351	0.8424

Framework versions

Transformers 4.23.1
Pytorch 1.12.1+cu113
Datasets 2.6.1
Tokenizers 0.13.1

Alex-VisTas/swin-tiny-patch4-window7-224-finetuned-woody_90epochs

swin-tiny-patch4-window7-224-finetuned-woody_90epochs

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions