review_classification_bert_base_jp_v3_ratio1_5_2label_add_dropout-epoch30_v3_test

This model is a fine-tuned version of cl-tohoku/bert-base-japanese-v3 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4153
• Accuracy: 0.8333

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	174	0.3584	0.8384
No log	2.0	348	0.3688	0.8384
0.3068	3.0	522	0.3905	0.8266
0.3068	4.0	696	0.4750	0.8384
0.3068	5.0	870	0.5981	0.8207
0.1234	6.0	1044	0.7474	0.8241
0.1234	7.0	1218	0.8165	0.8443
0.1234	8.0	1392	0.9306	0.8451
0.0219	9.0	1566	1.0272	0.8291
0.0219	10.0	1740	1.1183	0.8300
0.0219	11.0	1914	1.1176	0.8316
0.0038	12.0	2088	1.1603	0.8375
0.0038	13.0	2262	1.2502	0.8249
0.0038	14.0	2436	1.2183	0.8274
0.0018	15.0	2610	1.2581	0.8367
0.0018	16.0	2784	1.3006	0.8316
0.0018	17.0	2958	1.3206	0.8485
0.0004	18.0	3132	1.2802	0.8426
0.0004	19.0	3306	1.3309	0.8409
0.0004	20.0	3480	1.3235	0.8367
0.0022	21.0	3654	1.3566	0.8316
0.0022	22.0	3828	1.3597	0.8266
0.0011	23.0	4002	1.3692	0.8316
0.0011	24.0	4176	1.3597	0.8409
0.0011	25.0	4350	1.4123	0.8274
0.0001	26.0	4524	1.4911	0.8241
0.0001	27.0	4698	1.4161	0.8316
0.0001	28.0	4872	1.4187	0.8333
0.0001	29.0	5046	1.4226	0.8291
0.0001	30.0	5220	1.4153	0.8333

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0