2d3cc5298eb4be23a280b17b59222a20

This model is a fine-tuned version of google-bert/bert-large-cased on the contemmcm/hate-speech-and-offensive-language dataset. It achieves the following results on the evaluation set:

• Loss: 0.6851
• Data Size: 1.0
• Epoch Runtime: 68.1441
• Accuracy: 0.7672
• F1 Macro: 0.2894

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: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro
No log	0	0	1.4351	0	4.4494	0.1709	0.1108
No log	1	619	0.7291	0.0078	5.0852	0.7731	0.3158
No log	2	1238	0.6263	0.0156	6.1353	0.7672	0.2894
0.0151	3	1857	0.4239	0.0312	7.8471	0.8586	0.5419
0.0151	4	2476	0.3645	0.0625	9.4999	0.8825	0.5776
0.3356	5	3095	0.3005	0.125	13.3349	0.8892	0.6799
0.027	6	3714	0.2768	0.25	22.9399	0.9032	0.6088
0.2998	7	4333	0.3209	0.5	36.4790	0.8768	0.7435
0.2629	8.0	4952	0.3049	1.0	66.8236	0.9036	0.6223
0.6472	9.0	5571	0.6757	1.0	66.0054	0.7672	0.2894
0.6711	10.0	6190	0.6851	1.0	68.1441	0.7672	0.2894

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1