bert_chinese_mc_base-BioNER-EN-ZH

This model is a fine-tuned version of StivenLancheros/bert_chinese_mc_base-BioNER-EN on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3611
• Precision: 0.6967
• Recall: 0.7980
• F1: 0.7439
• Accuracy: 0.9215

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: 3e-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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.4895	1.0	680	0.6248	0.4389	0.6486	0.5235	0.8301
0.3569	2.0	1360	0.6207	0.4931	0.7204	0.5854	0.8481
0.2778	3.0	2040	0.4876	0.5723	0.7371	0.6443	0.8864
0.2558	4.0	2720	0.4496	0.5882	0.7446	0.6572	0.8892
0.2363	5.0	3400	0.4674	0.5845	0.7619	0.6615	0.8892
0.2129	6.0	4080	0.4311	0.6148	0.7674	0.6827	0.9005
0.2019	7.0	4760	0.3930	0.6428	0.7710	0.7011	0.9103
0.1912	8.0	5440	0.4031	0.6438	0.7815	0.7060	0.9095
0.1741	9.0	6120	0.3914	0.6506	0.7765	0.7080	0.9101
0.1727	10.0	6800	0.3808	0.6530	0.7814	0.7114	0.9117
0.1625	11.0	7480	0.4047	0.6545	0.7828	0.7129	0.9106
0.1546	12.0	8160	0.3803	0.6543	0.7849	0.7137	0.9115
0.1515	13.0	8840	0.3635	0.6828	0.7979	0.7359	0.9217
0.1415	14.0	9520	0.3872	0.6718	0.7962	0.7287	0.9160
0.1425	15.0	10200	0.3699	0.6879	0.7939	0.7371	0.9193
0.1327	16.0	10880	0.3762	0.6869	0.7977	0.7382	0.9184
0.1307	17.0	11560	0.3732	0.6822	0.8013	0.7369	0.9181
0.1309	18.0	12240	0.3629	0.6956	0.7970	0.7428	0.9208
0.1268	19.0	12920	0.3643	0.6930	0.7990	0.7423	0.9210
0.1257	20.0	13600	0.3611	0.6967	0.7980	0.7439	0.9215

Framework versions

• Transformers 4.27.2
• Pytorch 1.13.0+cu117
• Datasets 2.7.1
• Tokenizers 0.13.2