medical_bert_tiny_asymmetric_old_dataset_v2

This model is a fine-tuned version of prajjwal1/bert-tiny on an unknown dataset. It achieves the following results on the evaluation set:

Model description

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 0.0002
train_batch_size: 8
eval_batch_size: 64
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 16
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 0.1
num_epochs: 5
label_smoothing_factor: 0.1

Training Loss	Epoch	Step	Validation Loss	Accuracy	Recall Weighted	Precision Weighted	F1 Weighted	F1 Macro	F1 Prescription	F1 Lab Report	F1 Others	False Alarm Rate	Cross Class Error
0.0543	1.0	892	0.0200	0.9904	0.9904	0.9906	0.9904	0.9865	0.9784	0.9820	0.9991	0.0	0.0186
0.0011	2.0	1784	0.0095	0.9952	0.9952	0.9952	0.9952	0.9935	0.9910	0.9904	0.9991	0.0009	0.0083
0.0005	3.0	2676	0.0068	0.9962	0.9962	0.9962	0.9962	0.9949	0.9933	0.9923	0.9991	0.0	0.0062
0.0004	4.0	3568	0.0074	0.9962	0.9962	0.9962	0.9962	0.9949	0.9933	0.9923	0.9991	0.0	0.0062
0.0003	5.0	4460	0.0096	0.9952	0.9952	0.9952	0.9952	0.9931	0.9889	0.9904	1.0	0.0	0.0103