factual-consistency-multilabel-classification-ja

This model is a fine-tuned version of line-corporation/line-distilbert-base-japanese on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3230
• Accuracy: 0.8701

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: 8
• seed: 42
• distributed_type: tpu
• 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	306	0.4850	0.7861
0.5279	2.0	612	0.4657	0.7881
0.5279	3.0	918	0.4495	0.7979
0.4707	4.0	1224	0.4361	0.8115
0.448	5.0	1530	0.4252	0.8242
0.448	6.0	1836	0.4169	0.8271
0.4321	7.0	2142	0.4098	0.8369
0.4321	8.0	2448	0.4037	0.8389
0.4208	9.0	2754	0.3989	0.8428
0.4129	10.0	3060	0.3942	0.8447
0.4129	11.0	3366	0.3905	0.8457
0.4058	12.0	3672	0.3865	0.8467
0.4058	13.0	3978	0.3833	0.8496
0.4041	14.0	4284	0.3802	0.8496
0.3992	15.0	4590	0.3774	0.8506
0.3992	16.0	4896	0.3749	0.8525
0.3922	17.0	5202	0.3726	0.8525
0.3936	18.0	5508	0.3701	0.8535
0.3936	19.0	5814	0.3679	0.8535
0.3893	20.0	6120	0.3666	0.8535
0.3893	21.0	6426	0.3640	0.8555
0.3871	22.0	6732	0.3626	0.8545
0.3856	23.0	7038	0.3607	0.8555
0.3856	24.0	7344	0.3591	0.8564
0.3836	25.0	7650	0.3572	0.8584
0.3836	26.0	7956	0.3561	0.8604
0.3801	27.0	8262	0.3543	0.8604
0.3794	28.0	8568	0.3530	0.8613
0.3794	29.0	8874	0.3517	0.8633
0.379	30.0	9180	0.3505	0.8633
0.379	31.0	9486	0.3494	0.8633
0.377	32.0	9792	0.3482	0.8623
0.3765	33.0	10098	0.3471	0.8662
0.3765	34.0	10404	0.3465	0.8652
0.3739	35.0	10710	0.3456	0.8613
0.3737	36.0	11016	0.3441	0.8662
0.3737	37.0	11322	0.3435	0.8662
0.3723	38.0	11628	0.3426	0.8662
0.3723	39.0	11934	0.3418	0.8652
0.3728	40.0	12240	0.3410	0.8652
0.3713	41.0	12546	0.3401	0.8633
0.3713	42.0	12852	0.3395	0.8662
0.3686	43.0	13158	0.3392	0.8662
0.3686	44.0	13464	0.3378	0.8643
0.3693	45.0	13770	0.3375	0.8662
0.3685	46.0	14076	0.3365	0.8643
0.3685	47.0	14382	0.3362	0.8643
0.3675	48.0	14688	0.3352	0.8633
0.3675	49.0	14994	0.3349	0.8643
0.3654	50.0	15300	0.3341	0.8652
0.3672	51.0	15606	0.3339	0.8672
0.3672	52.0	15912	0.3335	0.8682
0.3659	53.0	16218	0.3325	0.8643
0.3648	54.0	16524	0.3323	0.8662
0.3648	55.0	16830	0.3315	0.8643
0.3638	56.0	17136	0.3314	0.8643
0.3638	57.0	17442	0.3307	0.8662
0.3657	58.0	17748	0.3304	0.8662
0.3641	59.0	18054	0.3299	0.8662
0.3641	60.0	18360	0.3297	0.8672
0.3624	61.0	18666	0.3294	0.8672
0.3624	62.0	18972	0.3290	0.8662
0.3625	63.0	19278	0.3285	0.8662
0.3639	64.0	19584	0.3281	0.8662
0.3639	65.0	19890	0.3282	0.8682
0.3632	66.0	20196	0.3274	0.8652
0.3618	67.0	20502	0.3273	0.8682
0.3618	68.0	20808	0.3270	0.8672
0.3636	69.0	21114	0.3267	0.8672
0.3636	70.0	21420	0.3265	0.8662
0.3577	71.0	21726	0.3262	0.8682
0.3607	72.0	22032	0.3262	0.8682
0.3607	73.0	22338	0.3258	0.8682
0.3591	74.0	22644	0.3255	0.8662
0.3591	75.0	22950	0.3255	0.8691
0.3597	76.0	23256	0.3252	0.8691
0.3593	77.0	23562	0.3250	0.8691
0.3593	78.0	23868	0.3248	0.8682
0.3597	79.0	24174	0.3246	0.8682
0.3597	80.0	24480	0.3244	0.8672
0.3593	81.0	24786	0.3243	0.8672
0.3602	82.0	25092	0.3242	0.8682
0.3602	83.0	25398	0.3242	0.8672
0.3579	84.0	25704	0.3242	0.8711
0.3614	85.0	26010	0.3238	0.8672
0.3614	86.0	26316	0.3238	0.8711
0.359	87.0	26622	0.3237	0.8701
0.359	88.0	26928	0.3236	0.8682
0.3583	89.0	27234	0.3235	0.8682
0.3591	90.0	27540	0.3234	0.8682
0.3591	91.0	27846	0.3233	0.8691
0.3581	92.0	28152	0.3231	0.8672
0.3581	93.0	28458	0.3232	0.8701
0.3579	94.0	28764	0.3231	0.8691
0.3584	95.0	29070	0.3231	0.8701
0.3584	96.0	29376	0.3230	0.8701
0.356	97.0	29682	0.3230	0.8691
0.356	98.0	29988	0.3230	0.8691
0.3604	99.0	30294	0.3230	0.8701
0.3607	100.0	30600	0.3230	0.8701

Framework versions

• Transformers 4.34.0
• Pytorch 2.0.0+cu118
• Datasets 2.14.5
• Tokenizers 0.14.0