scenario-kd-po-ner-full-xlmr-halfen_data-univner_en66

This model is a fine-tuned version of haryoaw/scenario-TCR-NER_data-univner_half on the None dataset. It achieves the following results on the evaluation set:

• Loss: 51.2856
• Precision: 0.7594
• Recall: 0.7774
• F1: 0.7683
• Accuracy: 0.9817

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: 66
• 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
84.3552	1.28	500	68.3345	0.7409	0.7164	0.7284	0.9800
61.0034	2.55	1000	61.0553	0.7614	0.7267	0.7436	0.9808
55.416	3.83	1500	57.0152	0.7573	0.7557	0.7565	0.9811
52.1923	5.1	2000	54.4194	0.7455	0.7671	0.7561	0.9813
50.0905	6.38	2500	52.9010	0.7549	0.7557	0.7553	0.9814
48.6873	7.65	3000	51.8337	0.7452	0.7598	0.7524	0.9815
47.8926	8.93	3500	51.2856	0.7594	0.7774	0.7683	0.9817

Framework versions

• Transformers 4.33.3
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3