clinical_trial_stop_reasons_custom

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

• Loss: 0.1448
• Accuracy Thresh: 0.9570
• F1 Micro: 0.5300
• F1 Macro: 0.1254
• Confusion Matrix: [[5940 15] [ 270 150]]

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 7

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy Thresh	F1 Micro	F1 Macro	Confusion Matrix
No log	1.0	106	0.2812	0.8328	0.0	0.0	[[5955 0]
[ 420 0]]
No log	2.0	212	0.2189	0.9382	0.0	0.0	[[5955 0]
[ 420 0]]
No log	3.0	318	0.1840	0.9489	0.0	0.0	[[5955 0]
[ 420 0]]
No log	4.0	424	0.1638	0.9485	0.4940	0.0989	[[5943 12]
[ 288 132]]
0.239	5.0	530	0.1526	0.9533	0.5060	0.1018	[[5943 12]
[ 277 143]]
0.239	6.0	636	0.1467	0.9564	0.5077	0.1020	[[5938 17]
[ 275 145]]
0.239	7.0	742	0.1448	0.9570	0.5300	0.1254	[[5940 15]
[ 270 150]]

Framework versions

• Transformers 4.26.0
• Pytorch 1.12.1+cu102
• Datasets 2.9.0
• Tokenizers 0.13.2