scenario-kd-pre-ner-full-mdeberta_data-univner_en55

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 62.9050
• Precision: 0.7596
• Recall: 0.7360
• F1: 0.7476
• Accuracy: 0.9801

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
148.0892	1.2755	500	103.3325	0.3059	0.2578	0.2798	0.9556
87.0932	2.5510	1000	80.2722	0.6704	0.6739	0.6722	0.9755
72.3221	3.8265	1500	72.3381	0.7265	0.7039	0.7150	0.9775
65.7687	5.1020	2000	68.3339	0.7549	0.7174	0.7357	0.9783
61.9669	6.3776	2500	65.6428	0.7442	0.7319	0.7380	0.9789
59.6427	7.6531	3000	64.0535	0.7581	0.7267	0.7421	0.9798
58.1252	8.9286	3500	62.9050	0.7596	0.7360	0.7476	0.9801

Framework versions

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