deit-base-distilled-patch16-224-hasta-85-fold2

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

• Loss: 1.0159
• Accuracy: 0.7273

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: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	1	1.3509	0.1818
No log	2.0	2	1.2094	0.1818
No log	3.0	3	1.0159	0.7273
No log	4.0	4	0.9047	0.6364
No log	5.0	5	0.9481	0.7273
No log	6.0	6	1.1337	0.7273
No log	7.0	7	1.2464	0.7273
No log	8.0	8	1.2335	0.7273
No log	9.0	9	1.1432	0.7273
0.4248	10.0	10	1.0979	0.7273
0.4248	11.0	11	1.1069	0.7273
0.4248	12.0	12	1.1633	0.7273
0.4248	13.0	13	1.1650	0.7273
0.4248	14.0	14	1.1332	0.7273
0.4248	15.0	15	1.1423	0.7273
0.4248	16.0	16	1.1208	0.7273
0.4248	17.0	17	1.0566	0.7273
0.4248	18.0	18	1.0348	0.7273
0.4248	19.0	19	1.0146	0.7273
0.1813	20.0	20	1.0062	0.7273
0.1813	21.0	21	1.0748	0.7273
0.1813	22.0	22	1.1673	0.7273
0.1813	23.0	23	1.2477	0.7273
0.1813	24.0	24	1.2457	0.7273
0.1813	25.0	25	1.1424	0.7273
0.1813	26.0	26	1.0439	0.7273
0.1813	27.0	27	0.9436	0.7273
0.1813	28.0	28	0.8254	0.7273
0.1813	29.0	29	0.7963	0.7273
0.1021	30.0	30	0.8055	0.7273
0.1021	31.0	31	0.7734	0.7273
0.1021	32.0	32	0.7069	0.7273
0.1021	33.0	33	0.7235	0.7273
0.1021	34.0	34	0.8200	0.7273
0.1021	35.0	35	0.8484	0.7273
0.1021	36.0	36	0.9380	0.7273
0.1021	37.0	37	1.0584	0.7273
0.1021	38.0	38	1.1803	0.7273
0.1021	39.0	39	1.2916	0.7273
0.0579	40.0	40	1.3442	0.7273
0.0579	41.0	41	1.3060	0.7273
0.0579	42.0	42	1.2349	0.7273
0.0579	43.0	43	1.1867	0.7273
0.0579	44.0	44	1.1411	0.7273
0.0579	45.0	45	1.2009	0.7273
0.0579	46.0	46	1.2366	0.7273
0.0579	47.0	47	1.2790	0.7273
0.0579	48.0	48	1.3573	0.7273
0.0579	49.0	49	1.4175	0.7273
0.0436	50.0	50	1.4130	0.7273
0.0436	51.0	51	1.4465	0.7273
0.0436	52.0	52	1.4372	0.7273
0.0436	53.0	53	1.4207	0.7273
0.0436	54.0	54	1.4134	0.7273
0.0436	55.0	55	1.4889	0.7273
0.0436	56.0	56	1.5351	0.7273
0.0436	57.0	57	1.5673	0.7273
0.0436	58.0	58	1.5803	0.7273
0.0436	59.0	59	1.5685	0.7273
0.0325	60.0	60	1.5487	0.7273
0.0325	61.0	61	1.4888	0.7273
0.0325	62.0	62	1.4026	0.7273
0.0325	63.0	63	1.2968	0.7273
0.0325	64.0	64	1.2915	0.7273
0.0325	65.0	65	1.3016	0.7273
0.0325	66.0	66	1.3871	0.7273
0.0325	67.0	67	1.5019	0.7273
0.0325	68.0	68	1.6563	0.7273
0.0325	69.0	69	1.8046	0.7273
0.028	70.0	70	1.9155	0.7273
0.028	71.0	71	1.9688	0.7273
0.028	72.0	72	1.9570	0.7273
0.028	73.0	73	1.8879	0.7273
0.028	74.0	74	1.8354	0.7273
0.028	75.0	75	1.7812	0.7273
0.028	76.0	76	1.6800	0.7273
0.028	77.0	77	1.5444	0.7273
0.028	78.0	78	1.4488	0.7273
0.028	79.0	79	1.3880	0.7273
0.0411	80.0	80	1.3544	0.7273
0.0411	81.0	81	1.3867	0.7273
0.0411	82.0	82	1.4348	0.7273
0.0411	83.0	83	1.4906	0.7273
0.0411	84.0	84	1.5685	0.7273
0.0411	85.0	85	1.6276	0.7273
0.0411	86.0	86	1.6793	0.7273
0.0411	87.0	87	1.7099	0.7273
0.0411	88.0	88	1.7182	0.7273
0.0411	89.0	89	1.7049	0.7273
0.0343	90.0	90	1.6831	0.7273
0.0343	91.0	91	1.6695	0.7273
0.0343	92.0	92	1.6584	0.7273
0.0343	93.0	93	1.6467	0.7273
0.0343	94.0	94	1.6294	0.7273
0.0343	95.0	95	1.6147	0.7273
0.0343	96.0	96	1.6055	0.7273
0.0343	97.0	97	1.6026	0.7273
0.0343	98.0	98	1.6043	0.7273
0.0343	99.0	99	1.6039	0.7273
0.021	100.0	100	1.6044	0.7273

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1