factual-consistency-classification-with-prompt-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: 0.6642
• Accuracy: 0.6738

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: 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	1.0023	0.3223
1.0128	2.0	612	0.9182	0.5742
1.0128	3.0	918	0.8847	0.5547
0.9226	4.0	1224	0.8569	0.5859
0.8815	5.0	1530	0.8329	0.6191
0.8815	6.0	1836	0.8287	0.5840
0.8633	7.0	2142	0.8160	0.5723
0.8633	8.0	2448	0.8210	0.5098
0.8525	9.0	2754	0.8171	0.5156
0.8418	10.0	3060	0.7850	0.5840
0.8418	11.0	3366	0.7771	0.6035
0.8323	12.0	3672	0.7652	0.6797
0.8323	13.0	3978	0.7655	0.6055
0.8292	14.0	4284	0.7556	0.6719
0.8233	15.0	4590	0.7578	0.6660
0.8233	16.0	4896	0.7497	0.6270
0.8173	17.0	5202	0.7472	0.6484
0.8122	18.0	5508	0.7334	0.7090
0.8122	19.0	5814	0.7468	0.6016
0.8165	20.0	6120	0.7248	0.7363
0.8165	21.0	6426	0.7324	0.6484
0.8048	22.0	6732	0.7261	0.6836
0.8033	23.0	7038	0.7187	0.7031
0.8033	24.0	7344	0.7205	0.6816
0.8011	25.0	7650	0.7307	0.6133
0.8011	26.0	7956	0.7220	0.6680
0.7979	27.0	8262	0.7175	0.6660
0.7963	28.0	8568	0.7205	0.6367
0.7963	29.0	8874	0.7139	0.6719
0.7942	30.0	9180	0.7078	0.6875
0.7942	31.0	9486	0.7094	0.6602
0.7878	32.0	9792	0.6992	0.7148
0.7891	33.0	10098	0.7041	0.6680
0.7891	34.0	10404	0.6968	0.6973
0.7869	35.0	10710	0.7047	0.6465
0.7874	36.0	11016	0.6962	0.6934
0.7874	37.0	11322	0.7026	0.6523
0.7817	38.0	11628	0.7103	0.625
0.7817	39.0	11934	0.6917	0.6914
0.7843	40.0	12240	0.6957	0.6680
0.7805	41.0	12546	0.7016	0.6484
0.7805	42.0	12852	0.6955	0.6582
0.7777	43.0	13158	0.7004	0.6387
0.7777	44.0	13464	0.6855	0.6895
0.7783	45.0	13770	0.6835	0.6895
0.7766	46.0	14076	0.6886	0.6641
0.7766	47.0	14382	0.6969	0.6309
0.7796	48.0	14688	0.6873	0.6738
0.7796	49.0	14994	0.6796	0.6953
0.77	50.0	15300	0.6908	0.6543
0.7768	51.0	15606	0.6900	0.6367
0.7768	52.0	15912	0.6855	0.6680
0.7698	53.0	16218	0.6905	0.6504
0.7686	54.0	16524	0.6783	0.6816
0.7686	55.0	16830	0.6807	0.6777
0.7712	56.0	17136	0.6767	0.6797
0.7712	57.0	17442	0.6966	0.6152
0.7692	58.0	17748	0.6812	0.6660
0.7677	59.0	18054	0.6762	0.6777
0.7677	60.0	18360	0.6697	0.7090
0.761	61.0	18666	0.6833	0.6445
0.761	62.0	18972	0.6753	0.6777
0.7676	63.0	19278	0.6757	0.6699
0.7627	64.0	19584	0.6874	0.6426
0.7627	65.0	19890	0.6704	0.6836
0.7672	66.0	20196	0.6685	0.6934
0.7638	67.0	20502	0.6645	0.7090
0.7638	68.0	20808	0.6718	0.6797
0.765	69.0	21114	0.6658	0.6934
0.765	70.0	21420	0.6670	0.6895
0.7593	71.0	21726	0.6735	0.6719
0.7634	72.0	22032	0.6765	0.6406
0.7634	73.0	22338	0.6722	0.6641
0.754	74.0	22644	0.6664	0.6855
0.754	75.0	22950	0.6659	0.6895
0.7619	76.0	23256	0.6700	0.6621
0.7583	77.0	23562	0.6664	0.6797
0.7583	78.0	23868	0.6650	0.6836
0.7556	79.0	24174	0.6615	0.6973
0.7556	80.0	24480	0.6625	0.6934
0.7571	81.0	24786	0.6704	0.6582
0.7549	82.0	25092	0.6677	0.6719
0.7549	83.0	25398	0.6670	0.6699
0.7542	84.0	25704	0.6617	0.6875
0.756	85.0	26010	0.6638	0.6758
0.756	86.0	26316	0.6697	0.6562
0.7513	87.0	26622	0.6647	0.6738
0.7513	88.0	26928	0.6734	0.6445
0.7548	89.0	27234	0.6637	0.6836
0.7565	90.0	27540	0.6665	0.6719
0.7565	91.0	27846	0.6708	0.6504
0.7488	92.0	28152	0.6603	0.6895
0.7488	93.0	28458	0.6671	0.6582
0.7545	94.0	28764	0.6655	0.6699
0.7509	95.0	29070	0.6636	0.6777
0.7509	96.0	29376	0.6620	0.6816
0.7546	97.0	29682	0.6653	0.6719
0.7546	98.0	29988	0.6636	0.6738
0.7521	99.0	30294	0.6636	0.6758
0.755	100.0	30600	0.6642	0.6738

Framework versions

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