7cf4f62bed321fdff5984be8f531b051

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

• Loss: 0.4784
• Data Size: 1.0
• Epoch Runtime: 18.2271
• Accuracy: 0.9081
• F1 Macro: 0.7288

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	1.1239	0	1.8287	0.1358	0.1343
No log	1	619	0.6859	0.0078	2.2907	0.7672	0.2894
No log	2	1238	0.6268	0.0156	2.2319	0.7672	0.2894
0.0145	3	1857	0.3729	0.0312	2.6538	0.8756	0.5694
0.0145	4	2476	0.3189	0.0625	3.0306	0.8916	0.5918
0.2893	5	3095	0.3008	0.125	4.1687	0.9048	0.7175
0.0239	6	3714	0.2752	0.25	6.3768	0.9103	0.6121
0.2582	7	4333	0.2598	0.5	10.3466	0.9079	0.7353
0.2278	8.0	4952	0.2447	1.0	18.4928	0.9138	0.7316
0.1467	9.0	5571	0.2965	1.0	18.5124	0.9099	0.7439
0.1125	10.0	6190	0.3383	1.0	18.2650	0.9034	0.7484
0.0855	11.0	6809	0.4312	1.0	18.3079	0.9056	0.7363
0.0452	12.0	7428	0.4784	1.0	18.2271	0.9081	0.7288

Framework versions

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