vineet1409/fine_tuned_bert_base_suicide_ideation_detection

results

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

• Loss: 0.2235
• Accuracy: 0.9579
• F1: 0.9579
• Precision: 0.9580
• Recall: 0.9579

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 10
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.1672	1.0	543	0.2845	0.9280	0.9279	0.9329	0.9285
0.2139	2.0	1086	0.2221	0.9439	0.9439	0.9447	0.9441
0.1531	3.0	1629	0.2010	0.9523	0.9523	0.9526	0.9522
0.0002	4.0	2172	0.2235	0.9579	0.9579	0.9580	0.9579

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Tokenizers 0.19.1