model_gpt2_d_hate_bias_gender_bias_ep_1_a_sqn_a_b_p_100_3_v_9

This model is a fine-tuned version of BogdanTurbal/model_gpt2_d_hate_bias_ep_1_sqn_a_p_100_v_9 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5347
• Balanced Accuracy: 0.7598
• Accuracy: 0.7617
• F1 Micro: 0.7617
• F1 Macro: 0.7579
• Precision Micro: 0.7617
• Precision Macro: 0.7746
• Recall Micro: 0.7617
• Recall Macro: 0.7598
• Auc: 0.8338

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Balanced Accuracy	Accuracy	F1 Micro	F1 Macro	Precision Micro	Precision Macro	Recall Micro	Recall Macro	Auc
0.8352	1.0	45	0.5972	0.6950	0.6965	0.6965	0.6936	0.6965	0.7014	0.6965	0.6950	0.7645
0.4604	2.0	90	0.5214	0.7469	0.7492	0.7492	0.7436	0.7492	0.7675	0.7492	0.7469	0.8211
0.395	3.0	135	0.5395	0.7542	0.7567	0.7567	0.7502	0.7567	0.7804	0.7567	0.7542	0.8303
0.3453	4.0	180	0.5347	0.7598	0.7617	0.7617	0.7579	0.7617	0.7746	0.7617	0.7598	0.8338

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1