scenario-kd-from-post-finetune-silver-div-2-data-hate_speech_filipino-model-hary

This model is a fine-tuned version of haryoaw/scenario-teacher-data-hate_speech_filipino-model-xlm-roberta-base on the hate_speech_filipino dataset. It achieves the following results on the evaluation set:

• Loss: 0.3428
• Accuracy: 0.7897
• F1: 0.7756

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 6969

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.32	100	0.8243	0.7597	0.7373
No log	0.64	200	0.7908	0.7448	0.7591
No log	0.96	300	0.5839	0.7698	0.7662
No log	1.28	400	0.5376	0.7708	0.7676
0.7939	1.6	500	0.5485	0.7788	0.7642
0.7939	1.92	600	0.5069	0.7701	0.7690
0.7939	2.24	700	0.5770	0.7821	0.7572
0.7939	2.56	800	0.4919	0.7772	0.7589
0.7939	2.88	900	0.4705	0.7760	0.7652
0.4564	3.19	1000	0.4889	0.7743	0.7677
0.4564	3.51	1100	0.5093	0.7859	0.7592
0.4564	3.83	1200	0.4266	0.7779	0.7666
0.4564	4.15	1300	0.5205	0.7838	0.7599
0.4564	4.47	1400	0.4029	0.7810	0.7681
0.3391	4.79	1500	0.4927	0.7786	0.7712
0.3391	5.11	1600	0.4180	0.7805	0.7642
0.3391	5.43	1700	0.4341	0.7772	0.7658
0.3391	5.75	1800	0.4398	0.7897	0.7635
0.3391	6.07	1900	0.3924	0.7791	0.7701
0.275	6.39	2000	0.3872	0.7828	0.7651
0.275	6.71	2100	0.4203	0.7807	0.7766
0.275	7.03	2200	0.3899	0.7828	0.7672
0.275	7.35	2300	0.3756	0.7805	0.7747
0.275	7.67	2400	0.3800	0.7736	0.7723
0.2453	7.99	2500	0.3945	0.7854	0.7714
0.2453	8.31	2600	0.3969	0.7762	0.7711
0.2453	8.63	2700	0.3726	0.7810	0.7737
0.2453	8.95	2800	0.3644	0.7836	0.7737
0.2453	9.27	2900	0.3730	0.7857	0.7793
0.2181	9.58	3000	0.3763	0.7850	0.7682
0.2181	9.9	3100	0.3599	0.7854	0.7775
0.2181	10.22	3200	0.3621	0.7909	0.7738
0.2181	10.54	3300	0.3554	0.7810	0.7714
0.2181	10.86	3400	0.3482	0.7885	0.7728
0.1999	11.18	3500	0.3554	0.7833	0.7751
0.1999	11.5	3600	0.3546	0.7840	0.7777
0.1999	11.82	3700	0.3986	0.7724	0.7674
0.1999	12.14	3800	0.3508	0.7857	0.7784
0.1999	12.46	3900	0.3365	0.7831	0.7732
0.1865	12.78	4000	0.3548	0.7810	0.7705
0.1865	13.1	4100	0.3534	0.7890	0.7764
0.1865	13.42	4200	0.3512	0.7810	0.7745
0.1865	13.74	4300	0.3747	0.7701	0.7719
0.1865	14.06	4400	0.3428	0.7897	0.7756

Framework versions

• Transformers 4.33.3
• Pytorch 2.0.1
• Datasets 2.14.5
• Tokenizers 0.13.3