40a70b3a3dd393543ce4086ee26b7d4f

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

• Loss: 0.6837
• Data Size: 1.0
• Epoch Runtime: 66.7474
• 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	1.5012	0	4.5637	0.0601	0.0378
No log	1	619	0.6886	0.0078	5.4289	0.7672	0.2894
No log	2	1238	0.6868	0.0156	5.9524	0.7672	0.2894
0.017	3	1857	0.5873	0.0312	7.7667	0.7672	0.2894
0.017	4	2476	0.5275	0.0625	9.6919	0.8703	0.5580
0.4396	5	3095	0.5119	0.125	12.8458	0.7672	0.2894
0.0542	6	3714	0.4946	0.25	20.3869	0.7672	0.2894
0.4426	7	4333	0.4583	0.5	35.0537	0.8267	0.5311
0.6799	8.0	4952	0.6810	1.0	67.5315	0.7672	0.2894
0.6441	9.0	5571	0.6784	1.0	66.0390	0.7672	0.2894
0.6653	10.0	6190	0.6886	1.0	67.0616	0.7672	0.2894
0.6606	11.0	6809	0.6837	1.0	66.7474	0.7672	0.2894

Framework versions

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