vit-base-cocoa

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

• Loss: 0.2061
• Accuracy: 0.9278

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: 8
• eval_batch_size: 8
• seed: 1337
• 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
• num_epochs: 100.0

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss
0.3733	1.0	196	0.9025	0.3558
0.3727	2.0	392	0.8989	0.4098
0.3901	3.0	588	0.8989	0.2668
0.3421	4.0	784	0.9170	0.2612
0.2703	5.0	980	0.9278	0.2061
0.1734	6.0	1176	0.9278	0.2568
0.1385	7.0	1372	0.9206	0.3242
0.3237	8.0	1568	0.9386	0.2922
0.236	9.0	1764	0.9386	0.3044
0.2124	10.0	1960	0.9061	0.3848
0.0454	11.0	2156	0.9350	0.3527
0.0756	12.0	2352	0.9350	0.2844
0.0605	13.0	2548	0.9314	0.3077
0.0214	14.0	2744	0.9025	0.6295
0.1816	15.0	2940	0.9386	0.2996
0.0338	16.0	3136	0.9278	0.3597
0.2136	17.0	3332	0.9314	0.4070
0.188	18.0	3528	0.9458	0.3532
0.0539	19.0	3724	0.9386	0.3843
0.0992	20.0	3920	0.9422	0.3904
0.0019	21.0	4116	0.9458	0.3732
0.0348	22.0	4312	0.9386	0.4021
0.0823	23.0	4508	0.9350	0.4217
0.1125	24.0	4704	0.9097	0.4704
0.0173	25.0	4900	0.9350	0.3700
0.0442	26.0	5096	0.9314	0.3725
0.0009	27.0	5292	0.9278	0.4819
0.0087	28.0	5488	0.9170	0.6492
0.0021	29.0	5684	0.9242	0.5297
0.2552	30.0	5880	0.9314	0.4482
0.0154	31.0	6076	0.9242	0.6075
0.0009	32.0	6272	0.9350	0.4101
0.1626	33.0	6468	0.9350	0.4653
0.0276	34.0	6664	0.9386	0.4174
0.0139	35.0	6860	0.9422	0.3992
0.0023	36.0	7056	0.9170	0.6972
0.1264	37.0	7252	0.9314	0.4980
0.0113	38.0	7448	0.9170	0.7154
0.0694	39.0	7644	0.9242	0.5443
0.0976	40.0	7840	0.9350	0.3852
0.1191	41.0	8036	0.9242	0.5398
0.1249	42.0	8232	0.9170	0.6197
0.0002	43.0	8428	0.9134	0.6967
0.1163	44.0	8624	0.9242	0.5697
0.0201	45.0	8820	0.9134	0.7221
0.0003	46.0	9016	0.9314	0.5253
0.0224	47.0	9212	0.9495	0.3817
0.0183	48.0	9408	0.9242	0.4966
0.0077	49.0	9604	0.9458	0.4349
0.0083	50.0	9800	0.9242	0.5191
0.0571	51.0	9996	0.9206	0.5826
0.0583	52.0	10192	0.9170	0.5335
0.0019	53.0	10388	0.9206	0.5843
0.0044	54.0	10584	0.9206	0.5895
0.0065	55.0	10780	0.9350	0.4487
0.0126	56.0	10976	0.9314	0.6221
0.0093	57.0	11172	0.9314	0.5138
0.0004	58.0	11368	0.9314	0.5162
0.0002	59.0	11564	0.9350	0.4514
0.1463	60.0	11760	0.9386	0.4744
0.0001	61.0	11956	0.9314	0.5338
0.0006	62.0	12152	0.9278	0.5788
0.0269	63.0	12348	0.9278	0.5500
0.1	64.0	12544	0.9206	0.6467
0.0004	65.0	12740	0.9242	0.5828
0.0001	66.0	12936	0.9314	0.5283
0.0001	67.0	13132	0.9206	0.6212
0.0002	68.0	13328	0.9242	0.4973
0.0058	69.0	13524	0.9278	0.5021
0.0605	70.0	13720	0.9170	0.6982
0.0006	71.0	13916	0.9350	0.4602
0.0021	72.0	14112	0.9314	0.5595
0.0004	73.0	14308	0.9386	0.4366
0.0124	74.0	14504	0.9134	0.7612
0.0284	75.0	14700	0.9206	0.6054
0.0001	76.0	14896	0.9242	0.5922
0.0119	77.0	15092	0.9242	0.5496
0.0006	78.0	15288	0.9206	0.6327
0.0711	79.0	15484	0.9386	0.5177
0.0001	80.0	15680	0.9134	0.7391
0.0985	81.0	15876	0.9242	0.5683
0.0001	82.0	16072	0.9206	0.6106
0.0	83.0	16268	0.9242	0.6235
0.0006	84.0	16464	0.9061	0.7914
0.0001	85.0	16660	0.9314	0.5649
0.0	86.0	16856	0.9350	0.5512
0.066	87.0	17052	0.9350	0.5473
0.0189	88.0	17248	0.9386	0.4866
0.0	89.0	17444	0.9386	0.5136
0.0001	90.0	17640	0.9350	0.5246
0.0001	91.0	17836	0.9314	0.5626
0.0037	92.0	18032	0.9350	0.5335
0.0999	93.0	18228	0.9242	0.6357
0.1124	94.0	18424	0.9278	0.5905
0.0175	95.0	18620	0.9206	0.6618
0.0001	96.0	18816	0.9386	0.5588
0.0259	97.0	19012	0.9350	0.5549
0.0001	98.0	19208	0.9350	0.5599
0.0285	99.0	19404	0.9350	0.5517
0.003	100.0	19600	0.9350	0.5570

Framework versions

• Transformers 4.48.0.dev0
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.21.0