deit-base-distilled-patch16-224-hasta-85-fold1

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: 0.9744
• 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	1.2772	0.0909
No log	2.0	2	1.1448	0.1818
No log	3.0	3	0.9744	0.7273
No log	4.0	4	0.9234	0.7273
No log	5.0	5	1.0760	0.7273
No log	6.0	6	1.3222	0.7273
No log	7.0	7	1.5248	0.7273
No log	8.0	8	1.6139	0.7273
No log	9.0	9	1.6711	0.7273
0.3554	10.0	10	1.7354	0.7273
0.3554	11.0	11	1.6721	0.7273
0.3554	12.0	12	1.5988	0.7273
0.3554	13.0	13	1.5960	0.7273
0.3554	14.0	14	1.5832	0.7273
0.3554	15.0	15	1.5683	0.7273
0.3554	16.0	16	1.5774	0.7273
0.3554	17.0	17	1.6468	0.7273
0.3554	18.0	18	1.7091	0.7273
0.3554	19.0	19	1.7276	0.7273
0.1335	20.0	20	1.7052	0.7273
0.1335	21.0	21	1.6426	0.7273
0.1335	22.0	22	1.5316	0.7273
0.1335	23.0	23	1.4017	0.7273
0.1335	24.0	24	1.3009	0.7273
0.1335	25.0	25	1.2864	0.7273
0.1335	26.0	26	1.3934	0.7273
0.1335	27.0	27	1.4436	0.7273
0.1335	28.0	28	1.5752	0.7273
0.1335	29.0	29	1.6211	0.7273
0.0769	30.0	30	1.5944	0.7273
0.0769	31.0	31	1.5283	0.7273
0.0769	32.0	32	1.4341	0.7273
0.0769	33.0	33	1.4512	0.7273
0.0769	34.0	34	1.4980	0.7273
0.0769	35.0	35	1.5803	0.7273
0.0769	36.0	36	1.7676	0.7273
0.0769	37.0	37	1.8581	0.7273
0.0769	38.0	38	1.8816	0.7273
0.0769	39.0	39	1.8317	0.7273
0.0505	40.0	40	1.7445	0.7273
0.0505	41.0	41	1.6965	0.7273
0.0505	42.0	42	1.7210	0.7273
0.0505	43.0	43	1.6903	0.7273
0.0505	44.0	44	1.6944	0.7273
0.0505	45.0	45	1.6923	0.7273
0.0505	46.0	46	1.7470	0.7273
0.0505	47.0	47	1.7502	0.7273
0.0505	48.0	48	1.7739	0.7273
0.0505	49.0	49	1.7819	0.7273
0.0255	50.0	50	1.8200	0.7273
0.0255	51.0	51	1.8122	0.7273
0.0255	52.0	52	1.7939	0.7273
0.0255	53.0	53	1.7736	0.7273
0.0255	54.0	54	1.7411	0.7273
0.0255	55.0	55	1.6773	0.7273
0.0255	56.0	56	1.6556	0.7273
0.0255	57.0	57	1.6767	0.7273
0.0255	58.0	58	1.6623	0.7273
0.0255	59.0	59	1.6553	0.7273
0.0227	60.0	60	1.6682	0.7273
0.0227	61.0	61	1.6209	0.7273
0.0227	62.0	62	1.6188	0.7273
0.0227	63.0	63	1.6919	0.7273
0.0227	64.0	64	1.7957	0.7273
0.0227	65.0	65	1.8750	0.7273
0.0227	66.0	66	1.9156	0.7273
0.0227	67.0	67	1.9163	0.7273
0.0227	68.0	68	1.8969	0.7273
0.0227	69.0	69	1.8814	0.7273
0.0185	70.0	70	1.8715	0.7273
0.0185	71.0	71	1.8892	0.7273
0.0185	72.0	72	1.9383	0.7273
0.0185	73.0	73	1.9627	0.7273
0.0185	74.0	74	2.0154	0.7273
0.0185	75.0	75	2.0326	0.7273
0.0185	76.0	76	2.0425	0.7273
0.0185	77.0	77	2.0586	0.7273
0.0185	78.0	78	2.0582	0.7273
0.0185	79.0	79	2.0863	0.7273
0.0246	80.0	80	2.1233	0.7273
0.0246	81.0	81	2.1527	0.7273
0.0246	82.0	82	2.1760	0.7273
0.0246	83.0	83	2.1907	0.7273
0.0246	84.0	84	2.1859	0.7273
0.0246	85.0	85	2.1654	0.7273
0.0246	86.0	86	2.1479	0.7273
0.0246	87.0	87	2.1194	0.7273
0.0246	88.0	88	2.1059	0.7273
0.0246	89.0	89	2.1032	0.7273
0.0228	90.0	90	2.0999	0.7273
0.0228	91.0	91	2.1037	0.7273
0.0228	92.0	92	2.1026	0.7273
0.0228	93.0	93	2.1132	0.7273
0.0228	94.0	94	2.1302	0.7273
0.0228	95.0	95	2.1453	0.7273
0.0228	96.0	96	2.1634	0.7273
0.0228	97.0	97	2.1762	0.7273
0.0228	98.0	98	2.1859	0.7273
0.0228	99.0	99	2.1916	0.7273
0.0142	100.0	100	2.1933	0.7273

Framework versions

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