cls-comment-phobert-base-v2-v2.1.2

This model is a fine-tuned version of vinai/phobert-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4647
• Accuracy: 0.9089
• F1 Score: 0.8724
• Recall: 0.8602
• Precision: 0.8863

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: 1e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 4000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Score	Recall	Precision
1.7278	2.02	100	1.5085	0.5057	0.1789	0.2120	0.1751
1.3435	4.04	200	1.0868	0.6516	0.2472	0.2862	0.2185
0.9504	6.06	300	0.7203	0.8	0.5338	0.5339	0.5372
0.6499	8.08	400	0.5300	0.8611	0.6612	0.6504	0.7187
0.4594	10.1	500	0.4458	0.8828	0.7132	0.6988	0.7379
0.347	12.12	600	0.4293	0.8764	0.7141	0.7140	0.7197
0.2744	14.14	700	0.3918	0.8866	0.7157	0.7208	0.7144
0.2174	16.16	800	0.3758	0.9013	0.8365	0.7997	0.8875
0.1608	18.18	900	0.3822	0.9	0.8552	0.8355	0.8792
0.126	20.2	1000	0.3968	0.9019	0.8614	0.8466	0.8831
0.1047	22.22	1100	0.4066	0.9006	0.8599	0.8487	0.8745
0.0891	24.24	1200	0.3824	0.9108	0.8696	0.8592	0.8837
0.0759	26.26	1300	0.3884	0.9185	0.8833	0.8640	0.9050
0.058	28.28	1400	0.4221	0.9070	0.8677	0.8573	0.8808
0.0495	30.3	1500	0.4308	0.9096	0.8683	0.8504	0.8897
0.044	32.32	1600	0.4303	0.9121	0.8680	0.8495	0.8908
0.0393	34.34	1700	0.4500	0.9083	0.8626	0.8461	0.8838
0.0333	36.36	1800	0.4647	0.9089	0.8724	0.8602	0.8863

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2