birds_transform_full

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Accuracy: 0.7303
• Loss: 1.4588

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss
5.6427	1.0	1984	0.4519	5.2504
4.6563	2.0	3968	0.5068	4.2749
3.6656	3.0	5952	0.5454	3.3311
2.7653	4.0	7936	0.5748	2.5181
2.0465	5.0	9920	0.6300	1.9205
1.5876	6.0	11904	0.6593	1.5696
1.3174	7.0	13888	0.6870	1.3831
1.1279	8.0	15872	0.7064	1.2516
1.0051	9.0	17856	0.7067	1.1999
0.9318	10.0	19840	0.7077	1.1631
0.8294	11.0	21824	0.7089	1.1444
0.7976	12.0	23808	0.7175	1.1156
0.7084	13.0	25792	0.7218	1.1209
0.6752	14.0	27776	0.7198	1.1032
0.6641	15.0	29760	0.7198	1.1192
0.6083	16.0	31744	0.7268	1.1044
0.5703	17.0	33728	0.7248	1.1287
0.5376	18.0	35712	0.7286	1.1115
0.5073	19.0	37696	0.7218	1.1429
0.5072	20.0	39680	0.7208	1.1519
0.4945	21.0	41664	0.7228	1.1636
0.4651	22.0	43648	0.7213	1.1771
0.4408	23.0	45632	0.7233	1.1650
0.4222	24.0	47616	0.7157	1.1841
0.409	25.0	49600	0.7145	1.2150
0.403	26.0	51584	0.7152	1.2203
0.3813	27.0	53568	0.7238	1.2064
0.3756	28.0	55552	0.7177	1.2526
0.365	29.0	57536	0.7208	1.2670
0.3729	30.0	59520	0.7180	1.2659
0.36	31.0	61504	0.7127	1.2545
0.3596	32.0	63488	0.7182	1.2728
0.3606	33.0	65472	0.7180	1.2886
0.325	34.0	67456	0.7157	1.2929
0.329	35.0	69440	0.7205	1.3074
0.3431	36.0	71424	0.7185	1.3122
0.3206	37.0	73408	0.7233	1.2993
0.3137	38.0	75392	0.7220	1.3206
0.3265	39.0	77376	0.7180	1.3246
0.3332	40.0	79360	0.7240	1.3163
0.3193	41.0	81344	0.7288	1.3259
0.3242	42.0	83328	0.7215	1.3320
0.2976	43.0	85312	0.7213	1.3283
0.3191	44.0	87296	0.7195	1.3453
0.3067	45.0	89280	0.7243	1.3550
0.2994	46.0	91264	0.7240	1.3324
0.3072	47.0	93248	0.7263	1.3412
0.2932	48.0	95232	0.7245	1.3345
0.2919	49.0	97216	0.7266	1.3759
0.2922	50.0	99200	0.7225	1.3873
0.304	51.0	101184	0.7235	1.3631
0.2898	52.0	103168	0.7205	1.3819
0.2773	53.0	105152	0.7251	1.3827
0.2756	54.0	107136	0.7228	1.3770
0.2789	55.0	109120	0.7248	1.3822
0.261	56.0	111104	0.7263	1.3878
0.2593	57.0	113088	0.7240	1.3955
0.2801	58.0	115072	0.7256	1.3659
0.2632	59.0	117056	0.7301	1.3719
0.2811	60.0	119040	0.7321	1.3775
0.2267	61.0	121024	0.7256	1.3689
0.2676	62.0	123008	0.7245	1.4069
0.2523	63.0	124992	0.7230	1.4166
0.2622	64.0	126976	0.7296	1.4018
0.2467	65.0	128960	0.7256	1.4287
0.2504	66.0	130944	0.7314	1.4019
0.2468	67.0	132928	0.7303	1.4058
0.2098	68.0	134912	0.7308	1.4093
0.2382	69.0	136896	0.7293	1.4206
0.2304	70.0	138880	0.7301	1.4078
0.251	71.0	140864	0.7251	1.4275
0.237	72.0	142848	0.7288	1.4283
0.2485	73.0	144832	0.7281	1.4338
0.2229	74.0	146816	0.7253	1.4386
0.2472	75.0	148800	0.7210	1.4440
0.2149	76.0	150784	0.7230	1.4319
0.2337	77.0	152768	0.7261	1.4422
0.2063	78.0	154752	0.7268	1.4456
0.216	79.0	156736	0.7218	1.4426
0.2249	80.0	158720	0.7198	1.4533
0.2148	81.0	160704	0.7230	1.4480
0.2321	82.0	162688	0.7273	1.4416
0.2306	83.0	164672	0.7286	1.4392
0.213	84.0	166656	0.7263	1.4609
0.2202	85.0	168640	0.7266	1.4590
0.206	86.0	170624	0.7245	1.4638
0.1987	87.0	172608	0.7251	1.4626
0.2181	88.0	174592	0.7261	1.4615
0.2076	89.0	176576	0.7253	1.4665
0.1999	90.0	178560	0.7251	1.4569
0.2287	91.0	180544	0.7266	1.4591
0.1985	92.0	182528	0.7263	1.4508
0.2166	93.0	184512	0.7266	1.4621
0.1943	94.0	186496	0.7276	1.4649
0.2189	95.0	188480	0.7293	1.4555
0.1911	96.0	190464	0.7306	1.4565
0.1954	97.0	192448	0.7271	1.4624
0.2053	98.0	194432	0.7286	1.4603
0.2067	99.0	196416	0.7306	1.4589
0.1917	100.0	198400	0.7303	1.4588

Framework versions

• Transformers 4.34.0
• Pytorch 2.1.0+cu121
• Datasets 2.14.5
• Tokenizers 0.14.1