finetuned-corrector

This model is a fine-tuned version of duyle2408/finetuned-corrector on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0040

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.0276	0.2133	5000	0.0142
0.017	0.4267	10000	0.0113
0.0137	0.6400	15000	0.0100
0.0119	0.8534	20000	0.0090
0.0097	1.0667	25000	0.0077
0.0074	1.2800	30000	0.0072
0.0069	1.4934	35000	0.0065
0.0063	1.7067	40000	0.0063
0.0057	1.9200	45000	0.0059
0.0046	2.1334	50000	0.0058
0.0039	2.3467	55000	0.0053
0.0036	2.5601	60000	0.0050
0.0034	2.7734	65000	0.0050
0.0032	2.9867	70000	0.0045
0.0022	3.2001	75000	0.0047
0.002	3.4134	80000	0.0047
0.0019	3.6267	85000	0.0044
0.0018	3.8401	90000	0.0042
0.0015	4.0534	95000	0.0043
0.001	4.2668	100000	0.0043
0.0009	4.4801	105000	0.0043
0.0008	4.6934	110000	0.0041
0.0008	4.9068	115000	0.0040

Framework versions

• Transformers 4.53.3
• Pytorch 2.7.1+cu126
• Datasets 3.6.0
• Tokenizers 0.21.4