bert-base-multilingual-cased-finetuned-ner

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: 0.2939
• Accuracy: 0.4501
• Precision: 0.5440
• Recall: 0.6659
• F1: 0.4954

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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	1.0	50	0.5390	0.3903	0.5183	0.4448	0.3118
No log	2.0	100	0.4150	0.4152	0.5575	0.5062	0.3632
No log	3.0	150	0.3530	0.4289	0.5842	0.5557	0.3945
No log	4.0	200	0.3272	0.4348	0.5319	0.5761	0.4145
No log	5.0	250	0.3047	0.4401	0.5175	0.6018	0.4284
No log	6.0	300	0.2964	0.4422	0.5224	0.6224	0.4600
No log	7.0	350	0.2927	0.4445	0.5391	0.6302	0.4691
No log	8.0	400	0.2896	0.4457	0.5295	0.6335	0.4668
No log	9.0	450	0.2810	0.4482	0.5360	0.6535	0.4846
0.324	10.0	500	0.2852	0.4486	0.5383	0.6554	0.4847
0.324	11.0	550	0.2949	0.4482	0.5372	0.6560	0.4858
0.324	12.0	600	0.2938	0.4494	0.5437	0.6603	0.4917
0.324	13.0	650	0.2906	0.4503	0.5437	0.6664	0.4952
0.324	14.0	700	0.2963	0.4499	0.5466	0.6641	0.4957
0.324	15.0	750	0.2939	0.4501	0.5440	0.6659	0.4954

Framework versions

• Transformers 4.40.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1