cls-comment-phobert-base-v2-v2.2.2

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: 0.6359
• Accuracy: 0.9235
• F1 Score: 0.8747
• Recall: 0.8709
• Precision: 0.8816

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
• label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Score	Recall	Precision
1.7287	1.05	100	1.5513	0.5041	0.1117	0.1667	0.0840
1.4006	2.11	200	1.1607	0.6608	0.3075	0.3196	0.4739
1.0635	3.16	300	0.8875	0.8212	0.5457	0.5578	0.5393
0.8514	4.21	400	0.7688	0.8522	0.5716	0.5872	0.5581
0.761	5.26	500	0.7055	0.8746	0.6412	0.6401	0.7368
0.6727	6.32	600	0.6545	0.9023	0.7811	0.7644	0.8581
0.6059	7.37	700	0.6360	0.9109	0.8464	0.8196	0.8859
0.5726	8.42	800	0.6340	0.9119	0.8564	0.8416	0.8743
0.5411	9.47	900	0.6197	0.9159	0.8692	0.8554	0.8868
0.5237	10.53	1000	0.6127	0.9192	0.8718	0.8474	0.9042
0.5055	11.58	1100	0.6201	0.9215	0.8703	0.8603	0.8839
0.5005	12.63	1200	0.6259	0.9231	0.8790	0.8680	0.8944
0.4846	13.68	1300	0.6159	0.9225	0.8726	0.8703	0.8759
0.4798	14.74	1400	0.6205	0.9244	0.8779	0.8636	0.8969
0.4744	15.79	1500	0.6254	0.9248	0.8742	0.8620	0.8909
0.4637	16.84	1600	0.6342	0.9228	0.8717	0.8653	0.8819
0.4584	17.89	1700	0.6359	0.9235	0.8747	0.8709	0.8816

Framework versions

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