theekshana/deid-roberta-i2b2-NER-medical-reports

fine-tuned-model

This model is a fine-tuned version of obi/deid_roberta_i2b2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2126
• Model Preparation Time: 0.0061
• Precision: 0.9143
• Recall: 0.9156
• F1: 0.9132
• Accuracy: 0.9156

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: 8
• seed: 42
• 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	Model Preparation Time	Precision	Recall	F1	Accuracy
0.4678	1.0	125	0.4248	0.0061	0.7737	0.8229	0.7836	0.8229
0.3877	2.0	250	0.4008	0.0061	0.7886	0.8282	0.8060	0.8282
0.3391	3.0	375	0.3132	0.0061	0.8213	0.8672	0.8389	0.8672
0.3091	4.0	500	0.3124	0.0061	0.8334	0.8597	0.8419	0.8597
0.2572	5.0	625	0.2570	0.0061	0.8675	0.8911	0.8739	0.8911
0.2368	6.0	750	0.2270	0.0061	0.8908	0.9084	0.8973	0.9084
0.2115	7.0	875	0.2219	0.0061	0.8960	0.9081	0.9017	0.9081
0.1949	8.0	1000	0.2325	0.0061	0.8993	0.9044	0.8991	0.9044
0.1843	9.0	1125	0.2218	0.0061	0.9035	0.9103	0.9059	0.9103
0.1691	10.0	1250	0.2126	0.0061	0.9143	0.9156	0.9132	0.9156

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu124
• Datasets 2.14.5
• Tokenizers 0.19.1