README.md

metadata

base_model: vinai/phobert-base-v2
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - recall
  - precision
model-index:
  - name: cls-comment-phobert-base-v2-v3.2.1
    results: []

cls-comment-phobert-base-v2-v3.2.1

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: 0.4657
• Accuracy: 0.9407
• F1 Score: 0.9319
• Recall: 0.9326
• Precision: 0.9319

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: 0.05

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Score	Recall	Precision
1.8571	0.8696	100	1.6996	0.3986	0.0814	0.1429	0.0569
1.552	1.7391	200	1.2878	0.6150	0.2552	0.2860	0.2738
1.1701	2.6087	300	0.9309	0.7746	0.5380	0.5249	0.5797
0.8958	3.4783	400	0.7468	0.8371	0.6099	0.6121	0.6113
0.7463	4.3478	500	0.6540	0.8641	0.6758	0.6741	0.7556
0.6489	5.2174	600	0.5884	0.8866	0.7502	0.7443	0.7611
0.5604	6.0870	700	0.5297	0.9010	0.8350	0.8196	0.9060
0.4907	6.9565	800	0.4928	0.9171	0.8962	0.8769	0.9190
0.4428	7.8261	900	0.4692	0.9220	0.9048	0.8958	0.9170
0.4086	8.6957	1000	0.4600	0.9236	0.9073	0.9183	0.8978
0.3892	9.5652	1100	0.4530	0.9293	0.9156	0.9159	0.9156
0.3659	10.4348	1200	0.4574	0.9258	0.9154	0.9258	0.9071
0.3577	11.3043	1300	0.4533	0.9288	0.9159	0.9177	0.9152
0.338	12.1739	1400	0.4454	0.9339	0.9203	0.9285	0.9128
0.3302	13.0435	1500	0.4539	0.9312	0.9179	0.9172	0.9196
0.3186	13.9130	1600	0.4533	0.9320	0.9220	0.9146	0.9298
0.3146	14.7826	1700	0.4485	0.9356	0.9246	0.9224	0.9281
0.3093	15.6522	1800	0.4557	0.9326	0.9194	0.9125	0.9291
0.3019	16.5217	1900	0.4684	0.9290	0.9169	0.9234	0.9128
0.2985	17.3913	2000	0.4545	0.9347	0.9248	0.9238	0.9259
0.2959	18.2609	2100	0.4689	0.9334	0.9220	0.9208	0.9249
0.2891	19.1304	2200	0.4558	0.9386	0.9262	0.9180	0.9360
0.2905	20.0	2300	0.4590	0.9358	0.9227	0.9163	0.9308
0.2875	20.8696	2400	0.4797	0.9307	0.9193	0.9146	0.9268
0.2812	21.7391	2500	0.4697	0.9356	0.9247	0.9257	0.9242
0.2789	22.6087	2600	0.4668	0.9380	0.9255	0.9250	0.9271
0.2785	23.4783	2700	0.4671	0.9383	0.9293	0.9301	0.9289
0.2773	24.3478	2800	0.4657	0.9391	0.9293	0.9274	0.9328
0.2814	25.2174	2900	0.4702	0.9361	0.9259	0.9285	0.9244
0.2744	26.0870	3000	0.4732	0.9353	0.9274	0.9290	0.9273
0.2772	26.9565	3100	0.4676	0.9388	0.9281	0.9301	0.9264
0.2736	27.8261	3200	0.4661	0.9394	0.9281	0.9242	0.9325
0.2754	28.6957	3300	0.4746	0.9367	0.9257	0.9233	0.9288
0.2717	29.5652	3400	0.4688	0.9380	0.9283	0.9255	0.9315
0.27	30.4348	3500	0.4697	0.9388	0.9304	0.9307	0.9308
0.2674	31.3043	3600	0.4668	0.9391	0.9274	0.9311	0.9240
0.2693	32.1739	3700	0.4657	0.9407	0.9319	0.9326	0.9319
0.2685	33.0435	3800	0.4672	0.9402	0.9298	0.9304	0.9297
0.268	33.9130	3900	0.4668	0.9410	0.9317	0.9311	0.9328
0.272	34.7826	4000	0.4654	0.9402	0.9310	0.9300	0.9325

Framework versions

• Transformers 4.40.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1