phobert_large_IS252

This model is a fine-tuned version of vinai/phobert-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2006
• Micro Precision: 0.9496
• Micro Recall: 0.9560
• Micro F1: 0.9528

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Micro Precision	Micro Recall	Micro F1
No log	1.0	79	0.2527	0.8824	0.9207	0.9011
No log	2.0	158	0.1526	0.9254	0.9591	0.9419
2.2763	3.0	237	0.1343	0.9422	0.9555	0.9488
2.2763	4.0	316	0.1480	0.9477	0.9597	0.9537
2.2763	5.0	395	0.1550	0.9431	0.9618	0.9523
0.2025	6.0	474	0.1852	0.9219	0.9428	0.9322
0.2025	7.0	553	0.1861	0.9505	0.9535	0.9520
0.1394	8.0	632	0.1750	0.9471	0.9624	0.9547
0.1394	9.0	711	0.1892	0.9556	0.9517	0.9536
0.1394	10.0	790	0.1690	0.9546	0.9616	0.9581
0.1294	11.0	869	0.2881	0.8426	0.8917	0.8664
0.1294	12.0	948	0.2265	0.9428	0.9563	0.9495
0.3906	13.0	1027	0.2446	0.9027	0.9476	0.9246
0.3906	14.0	1106	0.2606	0.8808	0.9548	0.9163
0.3906	15.0	1185	0.2564	0.8938	0.9592	0.9254

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.3
• Tokenizers 0.22.2