jaffe_V2_100_1

This model is a fine-tuned version of WinKawaks/vit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.6446
• Accuracy: 0.7333

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• 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	2.2977	0.1333
No log	2.0	2	2.0741	0.1
No log	3.0	3	2.0275	0.2333
No log	4.0	4	2.1234	0.0667
No log	5.0	5	2.0852	0.0333
No log	6.0	6	2.0259	0.1667
No log	7.0	7	2.0362	0.2667
No log	8.0	8	2.0153	0.2333
No log	9.0	9	1.7472	0.3333
1.9971	10.0	10	1.9598	0.2
1.9971	11.0	11	1.9367	0.3333
1.9971	12.0	12	1.8312	0.3333
1.9971	13.0	13	1.7299	0.3
1.9971	14.0	14	1.6306	0.4333
1.9971	15.0	15	1.5377	0.4333
1.9971	16.0	16	1.4326	0.5333
1.9971	17.0	17	1.5047	0.4
1.9971	18.0	18	1.4929	0.4333
1.9971	19.0	19	1.5326	0.4
1.5087	20.0	20	1.5017	0.4667
1.5087	21.0	21	1.4978	0.4333
1.5087	22.0	22	1.2678	0.5667
1.5087	23.0	23	1.2538	0.5
1.5087	24.0	24	1.2526	0.5
1.5087	25.0	25	1.3660	0.4
1.5087	26.0	26	1.3206	0.5
1.5087	27.0	27	1.2053	0.4333
1.5087	28.0	28	1.1457	0.6333
1.5087	29.0	29	1.0761	0.6
1.007	30.0	30	1.1556	0.5
1.007	31.0	31	1.0172	0.6333
1.007	32.0	32	1.1851	0.5333
1.007	33.0	33	1.0535	0.5333
1.007	34.0	34	1.1161	0.5333
1.007	35.0	35	0.9928	0.5667
1.007	36.0	36	0.9970	0.7
1.007	37.0	37	1.1090	0.4667
1.007	38.0	38	0.9536	0.6
1.007	39.0	39	1.2752	0.5
0.6664	40.0	40	0.8948	0.6667
0.6664	41.0	41	0.8891	0.6667
0.6664	42.0	42	0.8382	0.6333
0.6664	43.0	43	0.7498	0.7
0.6664	44.0	44	0.8668	0.6667
0.6664	45.0	45	1.1427	0.6667
0.6664	46.0	46	0.8066	0.6
0.6664	47.0	47	0.9161	0.6333
0.6664	48.0	48	0.8266	0.6
0.6664	49.0	49	0.9943	0.5667
0.469	50.0	50	0.6892	0.6333
0.469	51.0	51	0.7529	0.7667
0.469	52.0	52	0.9834	0.5333
0.469	53.0	53	0.8994	0.6
0.469	54.0	54	0.6394	0.7667
0.469	55.0	55	0.6854	0.7
0.469	56.0	56	0.6051	0.8
0.469	57.0	57	0.8493	0.6667
0.469	58.0	58	0.6897	0.7333
0.469	59.0	59	0.6698	0.6667
0.3604	60.0	60	0.6562	0.7667
0.3604	61.0	61	0.7638	0.6
0.3604	62.0	62	0.6217	0.7333
0.3604	63.0	63	0.7635	0.7
0.3604	64.0	64	0.7777	0.7667
0.3604	65.0	65	0.6505	0.8
0.3604	66.0	66	0.6469	0.7333
0.3604	67.0	67	0.7266	0.7333
0.3604	68.0	68	0.7613	0.6667
0.3604	69.0	69	0.4647	0.8
0.2726	70.0	70	0.6390	0.7
0.2726	71.0	71	0.6155	0.7333
0.2726	72.0	72	0.6113	0.8
0.2726	73.0	73	0.5648	0.8
0.2726	74.0	74	0.7042	0.7
0.2726	75.0	75	0.6263	0.8333
0.2726	76.0	76	0.7464	0.7333
0.2726	77.0	77	0.7640	0.6333
0.2726	78.0	78	0.7129	0.7667
0.2726	79.0	79	0.7362	0.7333
0.2157	80.0	80	0.7122	0.7667
0.2157	81.0	81	0.5565	0.7333
0.2157	82.0	82	0.6734	0.7667
0.2157	83.0	83	0.6057	0.7
0.2157	84.0	84	0.5287	0.7667
0.2157	85.0	85	0.7490	0.7333
0.2157	86.0	86	0.5841	0.7333
0.2157	87.0	87	0.5641	0.7667
0.2157	88.0	88	0.8243	0.6667
0.2157	89.0	89	0.5287	0.7667
0.1946	90.0	90	1.0455	0.7
0.1946	91.0	91	0.6091	0.7333
0.1946	92.0	92	0.5152	0.7667
0.1946	93.0	93	0.5850	0.8
0.1946	94.0	94	0.5806	0.7333
0.1946	95.0	95	0.6017	0.7667
0.1946	96.0	96	0.5606	0.7667
0.1946	97.0	97	0.5931	0.7667
0.1946	98.0	98	0.5299	0.7667
0.1946	99.0	99	0.7117	0.7333
0.1647	100.0	100	0.6446	0.7333

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.3.1
• Tokenizers 0.21.0