fa0e88ca2222ba812b586de129f72d47

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

• Loss: 0.6907
• Data Size: 1.0
• Epoch Runtime: 110.0629
• 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	1.1740	0	7.2551	0.0605	0.0381
No log	1	619	0.6779	0.0078	8.4279	0.7672	0.2894
No log	2	1238	0.7296	0.0156	9.9684	0.7672	0.2894
0.0166	3	1857	0.6821	0.0312	12.2328	0.7672	0.2894
0.0166	4	2476	0.6778	0.0625	15.8801	0.7672	0.2894
0.6631	5	3095	0.6874	0.125	23.1356	0.7672	0.2894
0.0642	6	3714	0.6774	0.25	36.1701	0.7672	0.2894
0.6684	7	4333	0.6786	0.5	61.5491	0.7672	0.2894
0.6756	8.0	4952	0.6810	1.0	109.8040	0.7672	0.2894
0.648	9.0	5571	0.6834	1.0	109.5082	0.7672	0.2894
0.6651	10.0	6190	0.6907	1.0	110.0629	0.7672	0.2894

Framework versions

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