disfluency-large-2

This model is a fine-tuned version of vinai/phobert-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0318
• Precision: 0.9837
• Recall: 0.9808
• F1: 0.9822
• Accuracy: 0.9946

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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	140	0.0439	0.9538	0.9561	0.9550	0.9890
No log	2.0	280	0.0314	0.9660	0.9736	0.9698	0.9906
No log	3.0	420	0.0394	0.9710	0.9651	0.9681	0.9909
0.1105	4.0	560	0.0320	0.9795	0.9784	0.9790	0.9929
0.1105	5.0	700	0.0450	0.9704	0.9657	0.9681	0.9904
0.1105	6.0	840	0.0463	0.9776	0.9694	0.9734	0.9911
0.1105	7.0	980	0.0480	0.9706	0.9712	0.9709	0.9909
0.0113	8.0	1120	0.0318	0.9837	0.9808	0.9822	0.9946
0.0113	9.0	1260	0.0419	0.9699	0.9669	0.9684	0.9915
0.0113	10.0	1400	0.0458	0.9735	0.9712	0.9723	0.9920
0.0051	11.0	1540	0.0309	0.9777	0.9766	0.9771	0.9935
0.0051	12.0	1680	0.0232	0.9820	0.9820	0.9820	0.9951
0.0051	13.0	1820	0.0344	0.9849	0.9784	0.9816	0.9945

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3