README.md

metadata

library_name: transformers
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: hate_speech
    results: []

hate_speech

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7687
• Accuracy: 0.8017
• Auc Score: 0.8728
• F1: 0.8298
• Precision: 0.8074
• Recall: 0.8534

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: 3e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Auc Score	F1	Precision	Recall
0.6445	0.0923	100	0.5441	0.7435	0.8052	0.7751	0.7701	0.7801
0.5767	0.1845	200	0.5260	0.7555	0.8345	0.7721	0.8179	0.7313
0.5126	0.2768	300	0.5090	0.7629	0.8450	0.8068	0.7493	0.8738
0.4723	0.3690	400	0.5557	0.7417	0.8505	0.7363	0.8728	0.6368
0.511	0.4613	500	0.4766	0.7823	0.8525	0.8106	0.7991	0.8225
0.5082	0.5535	600	0.4947	0.7915	0.8565	0.8239	0.7900	0.8607
0.4494	0.6458	700	0.4976	0.7763	0.8560	0.8032	0.8003	0.8062
0.4816	0.7380	800	0.4648	0.7827	0.8624	0.8111	0.7992	0.8233
0.4665	0.8303	900	0.4649	0.7887	0.8656	0.8200	0.7926	0.8493
0.5226	0.9225	1000	0.4537	0.7929	0.8666	0.8158	0.8222	0.8094
0.4643	1.0148	1100	0.4747	0.7998	0.8676	0.8287	0.8040	0.8550
0.3617	1.1070	1200	0.5402	0.7943	0.8668	0.8213	0.8084	0.8347
0.3439	1.1993	1300	0.5924	0.7966	0.8703	0.8267	0.7988	0.8567
0.3482	1.2915	1400	0.5369	0.8003	0.8681	0.8287	0.8060	0.8526
0.3855	1.3838	1500	0.5213	0.7966	0.8702	0.8205	0.8202	0.8208
0.335	1.4760	1600	0.5387	0.7929	0.8702	0.8176	0.8159	0.8192
0.382	1.5683	1700	0.5267	0.7924	0.8710	0.8109	0.8377	0.7858
0.341	1.6605	1800	0.6565	0.7957	0.8722	0.8293	0.7871	0.8762
0.3492	1.7528	1900	0.5635	0.7957	0.8725	0.8298	0.7855	0.8795
0.3861	1.8450	2000	0.5204	0.7998	0.8752	0.8281	0.8063	0.8510
0.3451	1.9373	2100	0.5854	0.7984	0.8757	0.8316	0.7893	0.8787
0.2915	2.0295	2200	0.6308	0.8021	0.8744	0.8354	0.7897	0.8868
0.2264	2.1218	2300	0.7711	0.7984	0.8741	0.8234	0.8172	0.8298
0.244	2.2140	2400	0.7302	0.8030	0.8742	0.8346	0.7960	0.8770
0.2477	2.3063	2500	0.8263	0.7915	0.8721	0.8154	0.8180	0.8127
0.2356	2.3985	2600	0.8275	0.7980	0.8734	0.8301	0.7926	0.8713
0.2122	2.4908	2700	0.8132	0.7980	0.8723	0.8234	0.8155	0.8314
0.2443	2.5830	2800	0.7874	0.8007	0.8728	0.8269	0.8139	0.8404
0.2275	2.6753	2900	0.7503	0.8003	0.8738	0.8322	0.7938	0.8746
0.2476	2.7675	3000	0.7822	0.7957	0.8731	0.8206	0.8163	0.8249
0.1961	2.8598	3100	0.7780	0.8021	0.8731	0.8304	0.8071	0.8550
0.2536	2.9520	3200	0.7687	0.8017	0.8728	0.8298	0.8074	0.8534

Framework versions

• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1