xlm-roberta-large-finetuned-ner-geocorpus

This model is a fine-tuned version of FacebookAI/xlm-roberta-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1447
• Precision: 0.8446
• Recall: 0.8970
• F1: 0.8700
• Accuracy: 0.9783

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: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.9991	275	0.1397	0.7230	0.7245	0.7237	0.9581
0.2086	1.9982	550	0.1079	0.7256	0.8507	0.7832	0.9655
0.2086	2.9973	825	0.0866	0.8121	0.8496	0.8304	0.9733
0.0636	4.0	1101	0.0830	0.8030	0.9001	0.8488	0.9759
0.0636	4.9991	1376	0.1024	0.8433	0.8770	0.8598	0.9756
0.032	5.9982	1651	0.1036	0.84	0.8833	0.8611	0.9784
0.032	6.9973	1926	0.1132	0.8278	0.8948	0.8600	0.9752
0.0165	8.0	2202	0.1211	0.8473	0.8927	0.8694	0.9782
0.0165	8.9991	2477	0.1406	0.8402	0.8959	0.8672	0.9775
0.0089	9.9909	2750	0.1447	0.8446	0.8970	0.8700	0.9783

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3