scenario-kd-pre-ner-full-xlmr_data-univner_en55

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

• Loss: 54.7950
• Precision: 0.7503
• Recall: 0.7526
• F1: 0.7514
• Accuracy: 0.9804

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: 3e-05
• train_batch_size: 8
• eval_batch_size: 32
• seed: 55
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
118.0158	1.2755	500	81.6998	0.7203	0.6398	0.6776	0.9757
69.7238	2.5510	1000	66.5568	0.7159	0.7277	0.7218	0.9783
60.1647	3.8265	1500	61.3497	0.7533	0.7112	0.7316	0.9795
55.862	5.1020	2000	58.5174	0.7655	0.7298	0.7472	0.9810
53.2325	6.3776	2500	56.8604	0.7608	0.7277	0.7439	0.9802
51.644	7.6531	3000	55.4373	0.7553	0.7443	0.7497	0.9808
50.5439	8.9286	3500	54.7950	0.7503	0.7526	0.7514	0.9804

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1