x5-ner-overfit-tuning

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

• Loss: 0.2588
• Precision: 0.9284
• Recall: 0.9505
• F1: 0.9394
• Accuracy: 0.9455

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: 8
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.8643	0.2754	500	0.3379	0.8295	0.8782	0.8532	0.9089
0.3556	0.5508	1000	0.3714	0.8562	0.8741	0.8650	0.9093
0.3058	0.8262	1500	0.2815	0.8730	0.9115	0.8919	0.9205
0.2563	1.1013	2000	0.2417	0.9085	0.9410	0.9244	0.9383
0.2188	1.3768	2500	0.2519	0.9209	0.9378	0.9293	0.9403
0.2102	1.6522	3000	0.2262	0.9230	0.9426	0.9327	0.9431
0.1851	1.9276	3500	0.2261	0.9237	0.9407	0.9321	0.9417
0.1558	2.2027	4000	0.2234	0.9290	0.9505	0.9396	0.9473
0.1339	2.4781	4500	0.2280	0.9297	0.9515	0.9404	0.9475
0.129	2.7535	5000	0.2588	0.9284	0.9505	0.9394	0.9455

Framework versions

• Transformers 4.53.3
• Pytorch 2.7.1+cu118
• Datasets 3.6.0
• Tokenizers 0.21.4