subcate-ts

This model is a fine-tuned version of tangminhanh/cate-ts on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0130
• Accuracy: 0.3377
• F1: 0.4963
• Precision: 0.9460
• Recall: 0.3364

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.8242	0.9987	188	0.2821	0.0	0.0081	0.0046	0.0321
0.1451	1.9973	376	0.0313	0.0	0.0	0.0	0.0
0.0298	2.9960	564	0.0231	0.0	0.0	0.0	0.0
0.024	4.0	753	0.0206	0.0	0.0	0.0	0.0
0.0212	4.9987	941	0.0181	0.1261	0.2233	0.9922	0.1258
0.0187	5.9973	1129	0.0162	0.2292	0.3696	0.9664	0.2285
0.0169	6.9960	1317	0.0147	0.3081	0.4657	0.9646	0.3070
0.0157	8.0	1506	0.0137	0.3184	0.4763	0.9551	0.3172
0.0151	8.9987	1694	0.0132	0.3310	0.4894	0.9486	0.3298
0.0145	9.9867	1880	0.0130	0.3377	0.4963	0.9460	0.3364

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1