factual-consistency-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: 0.6901
• Accuracy: 0.6855

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	1.0484	0.2773
1.044	2.0	612	1.0016	0.3320
1.044	3.0	918	0.9675	0.3633
0.9847	4.0	1224	0.9380	0.4434
0.9411	5.0	1530	0.9163	0.4863
0.9411	6.0	1836	0.8943	0.5742
0.9091	7.0	2142	0.8852	0.5137
0.9091	8.0	2448	0.8750	0.5039
0.8895	9.0	2754	0.8674	0.4941
0.8736	10.0	3060	0.8529	0.5254
0.8736	11.0	3366	0.8477	0.5195
0.8569	12.0	3672	0.8290	0.6133
0.8569	13.0	3978	0.8231	0.6055
0.8512	14.0	4284	0.8181	0.5879
0.8414	15.0	4590	0.8084	0.6211
0.8414	16.0	4896	0.8050	0.6152
0.8323	17.0	5202	0.8015	0.6016
0.8276	18.0	5508	0.7893	0.6426
0.8276	19.0	5814	0.7889	0.625
0.8248	20.0	6120	0.7829	0.6289
0.8248	21.0	6426	0.7796	0.6211
0.8166	22.0	6732	0.7759	0.6211
0.8107	23.0	7038	0.7720	0.6230
0.8107	24.0	7344	0.7691	0.625
0.8094	25.0	7650	0.7620	0.6426
0.8094	26.0	7956	0.7646	0.6230
0.8055	27.0	8262	0.7533	0.6582
0.8006	28.0	8568	0.7546	0.6348
0.8006	29.0	8874	0.7525	0.6348
0.7987	30.0	9180	0.7493	0.6465
0.7987	31.0	9486	0.7469	0.6484
0.7945	32.0	9792	0.7417	0.6562
0.7949	33.0	10098	0.7412	0.6465
0.7949	34.0	10404	0.7440	0.6367
0.7883	35.0	10710	0.7392	0.6504
0.7874	36.0	11016	0.7316	0.6660
0.7874	37.0	11322	0.7319	0.6543
0.7855	38.0	11628	0.7339	0.6504
0.7855	39.0	11934	0.7299	0.6562
0.7856	40.0	12240	0.7299	0.6504
0.7816	41.0	12546	0.7227	0.6738
0.7816	42.0	12852	0.7275	0.6504
0.7805	43.0	13158	0.7269	0.6543
0.7805	44.0	13464	0.7206	0.6641
0.7756	45.0	13770	0.7175	0.6777
0.7779	46.0	14076	0.7172	0.6660
0.7779	47.0	14382	0.7191	0.6582
0.7778	48.0	14688	0.7145	0.6680
0.7778	49.0	14994	0.7154	0.6602
0.7701	50.0	15300	0.7121	0.6719
0.774	51.0	15606	0.7142	0.6641
0.774	52.0	15912	0.7132	0.6719
0.7732	53.0	16218	0.7078	0.6836
0.768	54.0	16524	0.7123	0.6641
0.768	55.0	16830	0.7048	0.6855
0.7681	56.0	17136	0.7091	0.6641
0.7681	57.0	17442	0.7055	0.6797
0.7685	58.0	17748	0.7047	0.6816
0.7684	59.0	18054	0.7036	0.6836
0.7684	60.0	18360	0.7025	0.6836
0.7633	61.0	18666	0.7042	0.6699
0.7633	62.0	18972	0.7040	0.6699
0.7659	63.0	19278	0.7017	0.6777
0.7647	64.0	19584	0.7003	0.6836
0.7647	65.0	19890	0.7015	0.6738
0.7676	66.0	20196	0.6987	0.6816
0.7607	67.0	20502	0.6972	0.6875
0.7607	68.0	20808	0.6988	0.6777
0.7637	69.0	21114	0.6968	0.6875
0.7637	70.0	21420	0.6968	0.6816
0.7556	71.0	21726	0.6980	0.6738
0.7608	72.0	22032	0.6983	0.6758
0.7608	73.0	22338	0.6967	0.6758
0.7532	74.0	22644	0.6950	0.6816
0.7532	75.0	22950	0.6961	0.6738
0.7592	76.0	23256	0.6949	0.6797
0.7553	77.0	23562	0.6936	0.6836
0.7553	78.0	23868	0.6939	0.6855
0.7581	79.0	24174	0.6937	0.6816
0.7581	80.0	24480	0.6922	0.6836
0.7558	81.0	24786	0.6934	0.6758
0.7581	82.0	25092	0.6922	0.6855
0.7581	83.0	25398	0.6939	0.6738
0.7561	84.0	25704	0.6931	0.6797
0.7581	85.0	26010	0.6914	0.6836
0.7581	86.0	26316	0.6923	0.6797
0.7553	87.0	26622	0.6921	0.6816
0.7553	88.0	26928	0.6923	0.6797
0.7553	89.0	27234	0.6913	0.6816
0.7551	90.0	27540	0.6911	0.6816
0.7551	91.0	27846	0.6920	0.6777
0.7515	92.0	28152	0.6907	0.6797
0.7515	93.0	28458	0.6914	0.6797
0.7574	94.0	28764	0.6912	0.6797
0.7525	95.0	29070	0.6906	0.6836
0.7525	96.0	29376	0.6905	0.6836
0.7539	97.0	29682	0.6899	0.6855
0.7539	98.0	29988	0.6899	0.6855
0.754	99.0	30294	0.6901	0.6855
0.7573	100.0	30600	0.6901	0.6855

Framework versions

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