dit_maveriq_tobacco3482_2023-07-04_noaccum

This model is a fine-tuned version of microsoft/dit-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3767
• Accuracy: 0.95

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	25	1.6738	0.405
No log	2.0	50	1.2848	0.605
No log	3.0	75	0.9015	0.74
No log	4.0	100	0.6159	0.79
No log	5.0	125	0.4341	0.85
No log	6.0	150	0.3338	0.885
No log	7.0	175	0.3210	0.89
No log	8.0	200	0.3121	0.915
No log	9.0	225	0.3235	0.915
No log	10.0	250	0.3003	0.92
No log	11.0	275	0.2602	0.94
No log	12.0	300	0.2892	0.94
No log	13.0	325	0.2865	0.945
No log	14.0	350	0.3103	0.94
No log	15.0	375	0.2959	0.955
No log	16.0	400	0.3026	0.94
No log	17.0	425	0.3082	0.94
No log	18.0	450	0.2951	0.94
No log	19.0	475	0.3310	0.94
0.4279	20.0	500	0.3335	0.95
0.4279	21.0	525	0.3035	0.94
0.4279	22.0	550	0.3155	0.945
0.4279	23.0	575	0.3539	0.945
0.4279	24.0	600	0.3359	0.95
0.4279	25.0	625	0.3887	0.945
0.4279	26.0	650	0.3998	0.935
0.4279	27.0	675	0.4087	0.94
0.4279	28.0	700	0.4065	0.93
0.4279	29.0	725	0.3713	0.935
0.4279	30.0	750	0.3547	0.945
0.4279	31.0	775	0.3757	0.93
0.4279	32.0	800	0.3613	0.945
0.4279	33.0	825	0.3686	0.945
0.4279	34.0	850	0.3254	0.945
0.4279	35.0	875	0.3514	0.95
0.4279	36.0	900	0.3061	0.95
0.4279	37.0	925	0.3339	0.94
0.4279	38.0	950	0.3241	0.955
0.4279	39.0	975	0.2779	0.955
0.029	40.0	1000	0.2788	0.955
0.029	41.0	1025	0.2993	0.95
0.029	42.0	1050	0.3171	0.955
0.029	43.0	1075	0.3340	0.95
0.029	44.0	1100	0.3463	0.955
0.029	45.0	1125	0.3417	0.955
0.029	46.0	1150	0.3377	0.96
0.029	47.0	1175	0.3424	0.945
0.029	48.0	1200	0.3377	0.95
0.029	49.0	1225	0.3731	0.935
0.029	50.0	1250	0.3719	0.95
0.029	51.0	1275	0.3615	0.945
0.029	52.0	1300	0.3473	0.955
0.029	53.0	1325	0.3427	0.945
0.029	54.0	1350	0.4078	0.94
0.029	55.0	1375	0.3763	0.955
0.029	56.0	1400	0.3844	0.945
0.029	57.0	1425	0.3845	0.945
0.029	58.0	1450	0.3976	0.94
0.029	59.0	1475	0.3636	0.95
0.0115	60.0	1500	0.3431	0.95
0.0115	61.0	1525	0.3161	0.955
0.0115	62.0	1550	0.3482	0.945
0.0115	63.0	1575	0.3693	0.945
0.0115	64.0	1600	0.3435	0.95
0.0115	65.0	1625	0.3403	0.955
0.0115	66.0	1650	0.3644	0.95
0.0115	67.0	1675	0.3604	0.955
0.0115	68.0	1700	0.3746	0.945
0.0115	69.0	1725	0.3899	0.94
0.0115	70.0	1750	0.3684	0.95
0.0115	71.0	1775	0.4124	0.94
0.0115	72.0	1800	0.4010	0.95
0.0115	73.0	1825	0.3991	0.95
0.0115	74.0	1850	0.3859	0.95
0.0115	75.0	1875	0.3832	0.96
0.0115	76.0	1900	0.4054	0.955
0.0115	77.0	1925	0.4119	0.955
0.0115	78.0	1950	0.3724	0.955
0.0115	79.0	1975	0.3609	0.95
0.0116	80.0	2000	0.3663	0.955
0.0116	81.0	2025	0.3711	0.955
0.0116	82.0	2050	0.3730	0.955
0.0116	83.0	2075	0.3775	0.955
0.0116	84.0	2100	0.3805	0.96
0.0116	85.0	2125	0.3802	0.96
0.0116	86.0	2150	0.3773	0.96
0.0116	87.0	2175	0.3684	0.955
0.0116	88.0	2200	0.3750	0.95
0.0116	89.0	2225	0.3727	0.945
0.0116	90.0	2250	0.3742	0.945
0.0116	91.0	2275	0.3729	0.945
0.0116	92.0	2300	0.3727	0.945
0.0116	93.0	2325	0.3752	0.95
0.0116	94.0	2350	0.3726	0.945
0.0116	95.0	2375	0.3738	0.945
0.0116	96.0	2400	0.3747	0.945
0.0116	97.0	2425	0.3755	0.945
0.0116	98.0	2450	0.3757	0.95
0.0116	99.0	2475	0.3764	0.95
0.0061	100.0	2500	0.3767	0.95

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1.post200
• Datasets 2.9.0
• Tokenizers 0.13.2