Frozen11-BERT-multilingual-finetuned-CEFR_ner-3000news

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

• Loss: 0.5722
• Accuracy: 0.3776
• Precision: 0.5076
• Recall: 0.5639
• F1: 0.4282

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: 2e-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: 12

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	1.0	132	0.6172	0.3650	0.5084	0.5115	0.3886
No log	2.0	264	0.6107	0.3674	0.5113	0.5258	0.3962
No log	3.0	396	0.6047	0.3699	0.5045	0.5339	0.4050
0.5138	4.0	528	0.5961	0.3717	0.5108	0.5417	0.4106
0.5138	5.0	660	0.5889	0.3733	0.5119	0.5438	0.4156
0.5138	6.0	792	0.5854	0.3745	0.5108	0.5465	0.4169
0.5138	7.0	924	0.5809	0.3753	0.5043	0.5549	0.4204
0.4473	8.0	1056	0.5791	0.3762	0.5066	0.5546	0.4202
0.4473	9.0	1188	0.5773	0.3763	0.5034	0.5624	0.4245
0.4473	10.0	1320	0.5772	0.3765	0.5009	0.5625	0.4242
0.4473	11.0	1452	0.5724	0.3775	0.5049	0.5646	0.4278
0.4256	12.0	1584	0.5722	0.3776	0.5076	0.5639	0.4282

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1