434cc55a06243af6f166223cd3c81f92

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

• Loss: 0.6813
• Data Size: 1.0
• Epoch Runtime: 65.6256
• 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	1.8025	0	4.5920	0.0601	0.0378
No log	1	619	0.7092	0.0078	5.2692	0.7672	0.2894
No log	2	1238	0.6531	0.0156	5.9365	0.7672	0.2894
0.0161	3	1857	0.3981	0.0312	7.9195	0.875	0.5704
0.0161	4	2476	0.3884	0.0625	9.7741	0.8732	0.5715
0.4655	5	3095	0.3438	0.125	13.0385	0.8888	0.6652
0.0377	6	3714	0.3774	0.25	20.9520	0.8977	0.5987
0.6651	7	4333	0.6796	0.5	35.0415	0.7672	0.2894
0.6813	8.0	4952	0.6791	1.0	65.5053	0.7672	0.2894
0.6476	9.0	5571	0.6813	1.0	65.6256	0.7672	0.2894

Framework versions

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