vit-msn-small-beta-fia-manually-enhanced-HSV_test_4

This model is a fine-tuned version of Melo1512/vit-msn-small-beta-fia-manually-enhanced-HSV_test_3 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.3252
• Accuracy: 0.8662

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: 1e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• 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	0.3425	0.8803
No log	1.7143	3	0.3532	0.8803
No log	2.8571	5	0.3731	0.8732
No log	4.0	7	0.3582	0.8662
No log	4.5714	8	0.3560	0.8732
0.2214	5.7143	10	0.4090	0.8451
0.2214	6.8571	12	0.4253	0.8239
0.2214	8.0	14	0.3826	0.8592
0.2214	8.5714	15	0.3748	0.8592
0.2214	9.7143	17	0.3411	0.8592
0.2214	10.8571	19	0.3402	0.8521
0.1927	12.0	21	0.3833	0.8592
0.1927	12.5714	22	0.3901	0.8592
0.1927	13.7143	24	0.3608	0.8451
0.1927	14.8571	26	0.3565	0.8662
0.1927	16.0	28	0.3677	0.8803
0.1927	16.5714	29	0.3672	0.8732
0.213	17.7143	31	0.3429	0.8592
0.213	18.8571	33	0.3402	0.8803
0.213	20.0	35	0.3508	0.8662
0.213	20.5714	36	0.3578	0.8662
0.213	21.7143	38	0.3310	0.8662
0.1927	22.8571	40	0.3252	0.8662
0.1927	24.0	42	0.3473	0.8592
0.1927	24.5714	43	0.3671	0.8592
0.1927	25.7143	45	0.3863	0.8592
0.1927	26.8571	47	0.3622	0.8592
0.1927	28.0	49	0.3521	0.8592
0.1856	28.5714	50	0.3529	0.8592
0.1856	29.7143	52	0.3596	0.8592
0.1856	30.8571	54	0.3648	0.8592
0.1856	32.0	56	0.3637	0.8662
0.1856	32.5714	57	0.3686	0.8592
0.1856	33.7143	59	0.3602	0.8521
0.18	34.8571	61	0.3648	0.8662
0.18	36.0	63	0.3529	0.8521
0.18	36.5714	64	0.3595	0.8662
0.18	37.7143	66	0.3939	0.8521
0.18	38.8571	68	0.4526	0.8239
0.2048	40.0	70	0.4505	0.8239
0.2048	40.5714	71	0.4319	0.8451
0.2048	41.7143	73	0.3779	0.8521
0.2048	42.8571	75	0.3352	0.8592
0.2048	44.0	77	0.3353	0.8662
0.2048	44.5714	78	0.3450	0.8521
0.1977	45.7143	80	0.3557	0.8521
0.1977	46.8571	82	0.3715	0.8521
0.1977	48.0	84	0.3852	0.8521
0.1977	48.5714	85	0.3962	0.8521
0.1977	49.7143	87	0.4063	0.8521
0.1977	50.8571	89	0.4012	0.8521
0.1952	52.0	91	0.3912	0.8521
0.1952	52.5714	92	0.3874	0.8521
0.1952	53.7143	94	0.3822	0.8521
0.1952	54.8571	96	0.3758	0.8592
0.1952	56.0	98	0.3722	0.8592
0.1952	56.5714	99	0.3720	0.8592
0.1714	57.1429	100	0.3717	0.8592

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.2.0
• Tokenizers 0.19.1