cls-comment-phobert-base-v2-v2.2

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: 0.3163
• Accuracy: 0.9317
• F1 Score: 0.8875
• Recall: 0.8777
• Precision: 0.9015

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
1.663	1.05	100	1.4870	0.5041	0.1117	0.1667	0.0840
1.294	2.11	200	0.9956	0.6975	0.3911	0.3900	0.4902
0.898	3.16	300	0.6779	0.8232	0.5499	0.5697	0.5354
0.6411	4.21	400	0.5164	0.8568	0.5740	0.5895	0.5613
0.5031	5.26	500	0.4106	0.8938	0.7181	0.7114	0.7319
0.38	6.32	600	0.3474	0.9096	0.8326	0.8145	0.8739
0.2927	7.37	700	0.3110	0.9142	0.8598	0.8455	0.8810
0.2532	8.42	800	0.3046	0.9188	0.8702	0.8551	0.8881
0.2049	9.47	900	0.2851	0.9218	0.8689	0.8539	0.8902
0.1785	10.53	1000	0.2802	0.9251	0.8769	0.8561	0.9045
0.1511	11.58	1100	0.2875	0.9231	0.8744	0.8748	0.8770
0.1392	12.63	1200	0.2811	0.9264	0.8775	0.8597	0.9005
0.1166	13.68	1300	0.2757	0.9248	0.8751	0.8746	0.8786
0.1087	14.74	1400	0.2727	0.9258	0.8804	0.8761	0.8858
0.0918	15.79	1500	0.2862	0.9284	0.8830	0.8712	0.8988
0.0824	16.84	1600	0.2915	0.9291	0.8833	0.8689	0.9009
0.0745	17.89	1700	0.2994	0.9291	0.8797	0.8796	0.8847
0.0743	18.95	1800	0.3092	0.9254	0.8783	0.8686	0.8910
0.0636	20.0	1900	0.3142	0.9291	0.8811	0.8743	0.8916
0.0605	21.05	2000	0.3099	0.9291	0.8823	0.8700	0.8974
0.0501	22.11	2100	0.3163	0.9317	0.8875	0.8777	0.9015
0.0519	23.16	2200	0.3290	0.9297	0.8837	0.8692	0.9011
0.0464	24.21	2300	0.3406	0.9274	0.8805	0.8772	0.8872
0.0432	25.26	2400	0.3305	0.9284	0.8810	0.8775	0.8876
0.0404	26.32	2500	0.3378	0.9294	0.8826	0.8785	0.8901
0.0416	27.37	2600	0.3436	0.9284	0.8830	0.8726	0.8977

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2