scenario-kd-po-ner-full-mdeberta-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: 63.9106
• Precision: 0.7647
• Recall: 0.7671
• F1: 0.7659
• Accuracy: 0.9812

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
115.4337	1.28	500	89.9508	0.6039	0.5114	0.5538	0.9672
81.5056	2.55	1000	78.2589	0.7137	0.7019	0.7077	0.9781
73.2299	3.83	1500	72.9013	0.7153	0.7360	0.7255	0.9791
68.3008	5.1	2000	69.0462	0.7313	0.7692	0.7497	0.9806
64.8932	6.38	2500	66.5653	0.7296	0.7516	0.7404	0.9796
62.7315	7.65	3000	64.7790	0.7677	0.7526	0.7601	0.9808
61.2637	8.93	3500	63.9106	0.7647	0.7671	0.7659	0.9812

Framework versions

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