bert-finetuned-MedicalChunk

This model is a fine-tuned version of bert-base-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1762
• Precision: 0.2723
• Recall: 0.3065
• F1: 0.2884
• Accuracy: 0.9563

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	Precision	Recall	F1	Accuracy
No log	1.0	56	0.1631	0.0	0.0	0.0	0.9606
No log	2.0	112	0.1416	0.0638	0.0302	0.0410	0.9592
No log	3.0	168	0.1405	0.1982	0.2161	0.2067	0.9559
No log	4.0	224	0.1356	0.2771	0.2312	0.2521	0.9633
No log	5.0	280	0.1419	0.2928	0.2663	0.2789	0.9593
No log	6.0	336	0.1550	0.2732	0.2513	0.2618	0.9602
No log	7.0	392	0.1620	0.2732	0.2814	0.2772	0.9578
No log	8.0	448	0.1670	0.2585	0.3065	0.2805	0.9554
0.1137	9.0	504	0.1728	0.2553	0.3015	0.2765	0.9552
0.1137	10.0	560	0.1762	0.2723	0.3065	0.2884	0.9563

Framework versions

• Transformers 4.22.1
• Pytorch 1.12.1
• Datasets 2.4.0
• Tokenizers 0.12.1