deit-base-distilled-patch16-224-hasta-55-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: 1.0232
• Accuracy: 0.6667

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	0.5714	1	1.3241	0.3611
No log	1.7143	3	1.1598	0.3056
No log	2.8571	5	1.1928	0.3889
No log	4.0	7	1.1229	0.3333
No log	4.5714	8	1.1111	0.3333
1.13	5.7143	10	1.1062	0.3611
1.13	6.8571	12	1.0655	0.4722
1.13	8.0	14	1.0527	0.3889
1.13	8.5714	15	1.0448	0.4444
1.13	9.7143	17	0.9955	0.5556
1.13	10.8571	19	0.9819	0.5556
0.9629	12.0	21	0.9722	0.5
0.9629	12.5714	22	0.9777	0.4444
0.9629	13.7143	24	0.9126	0.5556
0.9629	14.8571	26	0.9932	0.5556
0.9629	16.0	28	0.9946	0.5833
0.9629	16.5714	29	0.9690	0.5556
0.822	17.7143	31	0.9163	0.5833
0.822	18.8571	33	0.9799	0.6389
0.822	20.0	35	0.9273	0.6389
0.822	20.5714	36	0.9008	0.6389
0.822	21.7143	38	1.0305	0.5
0.7066	22.8571	40	0.9330	0.5833
0.7066	24.0	42	0.9151	0.5833
0.7066	24.5714	43	0.9214	0.6111
0.7066	25.7143	45	0.9322	0.6111
0.7066	26.8571	47	1.0534	0.5833
0.7066	28.0	49	1.1223	0.5556
0.5702	28.5714	50	1.0081	0.5833
0.5702	29.7143	52	0.8680	0.6389
0.5702	30.8571	54	0.9259	0.6111
0.5702	32.0	56	0.9936	0.6111
0.5702	32.5714	57	0.9762	0.6111
0.5702	33.7143	59	0.9298	0.6111
0.4903	34.8571	61	0.9352	0.6111
0.4903	36.0	63	0.9919	0.5833
0.4903	36.5714	64	0.9661	0.5833
0.4903	37.7143	66	0.9764	0.6389
0.4903	38.8571	68	0.9909	0.6389
0.3959	40.0	70	0.9906	0.6389
0.3959	40.5714	71	0.9789	0.6389
0.3959	41.7143	73	0.9529	0.6389
0.3959	42.8571	75	0.9729	0.6389
0.3959	44.0	77	1.0425	0.6111
0.3959	44.5714	78	1.0531	0.6111
0.3792	45.7143	80	0.9800	0.5833
0.3792	46.8571	82	0.9609	0.5833
0.3792	48.0	84	0.9820	0.5833
0.3792	48.5714	85	1.0252	0.6389
0.3792	49.7143	87	1.0592	0.6111
0.3792	50.8571	89	1.0232	0.6667
0.3267	52.0	91	0.9972	0.6111
0.3267	52.5714	92	0.9925	0.6111
0.3267	53.7143	94	1.0007	0.5833
0.3267	54.8571	96	1.0169	0.5833
0.3267	56.0	98	1.0304	0.5833
0.3267	56.5714	99	1.0360	0.6111
0.2959	57.1429	100	1.0385	0.6111

Framework versions

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