xlm-roberta-base-NER-favsbot

This model is a fine-tuned version of xlm-roberta-base on the favsbot dataset. It achieves the following results on the evaluation set:

• Loss: 1.0572
• Precision: 0.5556
• Recall: 0.4722
• F1: 0.5105
• Accuracy: 0.6900

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: 1.5e-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
No log	1.0	4	2.4303	0.1448	0.3556	0.2058	0.1855
No log	2.0	8	2.3220	0.1465	0.3556	0.2075	0.1991
No log	3.0	12	2.1842	0.2486	0.2389	0.2436	0.4593
No log	4.0	16	1.9552	0.4	0.0111	0.0216	0.4367
No log	5.0	20	1.6989	0.0	0.0	0.0	0.4321
No log	6.0	24	1.6532	0.5	0.0056	0.0110	0.4344
No log	7.0	28	1.5724	0.3649	0.15	0.2126	0.5045
No log	8.0	32	1.5164	0.3654	0.2111	0.2676	0.5271
No log	9.0	36	1.4448	0.4203	0.1611	0.2329	0.5090
No log	10.0	40	1.3922	0.4833	0.1611	0.2417	0.5158
No log	11.0	44	1.3409	0.5395	0.2278	0.3203	0.5498
No log	12.0	48	1.2831	0.5824	0.2944	0.3911	0.5950
No log	13.0	52	1.2269	0.5714	0.3556	0.4384	0.6335
No log	14.0	56	1.1766	0.5625	0.4	0.4675	0.6606
No log	15.0	60	1.1408	0.5540	0.4278	0.4828	0.6674
No log	16.0	64	1.1159	0.56	0.4667	0.5091	0.6810
No log	17.0	68	1.0908	0.5658	0.4778	0.5181	0.6855
No log	18.0	72	1.0722	0.5658	0.4778	0.5181	0.6923
No log	19.0	76	1.0615	0.5592	0.4722	0.5120	0.6900
No log	20.0	80	1.0572	0.5556	0.4722	0.5105	0.6900

Framework versions

• Transformers 4.21.1
• Pytorch 1.12.1
• Datasets 2.4.0
• Tokenizers 0.12.1