2fd93761cee38dd43e0ddc98af316a8f

This model is a fine-tuned version of albert/albert-large-v2 on the contemmcm/hate-speech-and-offensive-language dataset. It achieves the following results on the evaluation set:

• Loss: 0.6704
• Data Size: 1.0
• Epoch Runtime: 43.9994
• 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	1.1244	0	3.8580	0.1670	0.1314
No log	1	619	0.7125	0.0078	4.5480	0.7672	0.2894
No log	2	1238	0.6460	0.0156	4.5558	0.7672	0.2894
0.016	3	1857	0.5565	0.0312	5.1657	0.7987	0.4105
0.016	4	2476	0.6316	0.0625	7.4823	0.7672	0.2894
0.6005	5	3095	0.4552	0.125	8.9755	0.8521	0.5452
0.0384	6	3714	0.4566	0.25	14.2178	0.8722	0.5664
0.6126	7	4333	0.5186	0.5	24.7002	0.8462	0.5343
0.6776	8.0	4952	0.6786	1.0	44.3803	0.7672	0.2894
0.644	9.0	5571	0.6704	1.0	43.9994	0.7672	0.2894

Framework versions

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