multibert1110_lrate10b16

This model is a fine-tuned version of bert-base-multilingual-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5469
• Precisions: 0.8548
• Recall: 0.8081
• F-measure: 0.8287
• Accuracy: 0.9073

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: 0.0001
• 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: 14

Training results

Training Loss	Epoch	Step	Validation Loss	Precisions	Recall	F-measure	Accuracy
0.6155	1.0	236	0.4117	0.8520	0.6645	0.6887	0.8660
0.3587	2.0	472	0.3608	0.7877	0.7387	0.7562	0.8864
0.2315	3.0	708	0.3620	0.8962	0.7550	0.7918	0.8977
0.1551	4.0	944	0.4640	0.8523	0.7478	0.7834	0.8931
0.1117	5.0	1180	0.4567	0.8269	0.7425	0.7712	0.8958
0.0885	6.0	1416	0.4916	0.8679	0.7882	0.8206	0.9037
0.0646	7.0	1652	0.5469	0.8548	0.8081	0.8287	0.9073
0.0385	8.0	1888	0.5638	0.8665	0.7813	0.8064	0.8999
0.0262	9.0	2124	0.5864	0.8872	0.7415	0.7881	0.9045
0.0231	10.0	2360	0.5984	0.8577	0.7582	0.7966	0.9017
0.0114	11.0	2596	0.6513	0.8594	0.7532	0.7930	0.9032
0.0119	12.0	2832	0.6270	0.8717	0.7591	0.8007	0.9034
0.006	13.0	3068	0.6814	0.8733	0.7411	0.7864	0.9034
0.0041	14.0	3304	0.6782	0.8722	0.7505	0.7943	0.9040

Framework versions

• Transformers 4.34.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.14.1