8a353c716e8ba6e3c095e28d57a1efb7

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

• Loss: 0.6758
• Data Size: 1.0
• Epoch Runtime: 65.4337
• 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	0.9945	0	4.3310	0.6236	0.3218
No log	1	619	0.6693	0.0078	5.2095	0.7672	0.2894
No log	2	1238	0.6252	0.0156	5.6753	0.7672	0.2894
0.0155	3	1857	0.4925	0.0312	7.9543	0.8196	0.4930
0.0155	4	2476	0.3797	0.0625	9.7324	0.9004	0.6004
0.3749	5	3095	0.3327	0.125	13.4013	0.9024	0.6020
0.0443	6	3714	0.6882	0.25	21.9062	0.7672	0.2894
0.661	7	4333	0.6777	0.5	35.4167	0.7672	0.2894
0.6812	8.0	4952	0.6806	1.0	65.8175	0.7672	0.2894
0.6459	9.0	5571	0.6758	1.0	65.4337	0.7672	0.2894

Framework versions

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