c220c2415a4bc9ffefd518a22abfadf2

This model is a fine-tuned version of FacebookAI/xlm-roberta-large-finetuned-conll03-english on the contemmcm/hate-speech-and-offensive-language dataset. It achieves the following results on the evaluation set:

• Loss: 0.6835
• Data Size: 1.0
• Epoch Runtime: 106.5929
• 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	1.0559	0	7.4227	0.4166	0.2885
No log	1	619	0.6986	0.0078	8.6091	0.7672	0.2894
No log	2	1238	0.6531	0.0156	11.2017	0.7672	0.2894
0.0155	3	1857	0.5238	0.0312	12.3328	0.7871	0.3625
0.0155	4	2476	0.4166	0.0625	15.7341	0.8941	0.5915
0.3436	5	3095	0.3045	0.125	22.4207	0.8981	0.7303
0.0647	6	3714	0.6792	0.25	35.7145	0.7672	0.2894
0.6022	7	4333	0.6803	0.5	61.0734	0.7672	0.2894
0.676	8.0	4952	0.6805	1.0	107.0124	0.7672	0.2894
0.6491	9.0	5571	0.6835	1.0	106.5929	0.7672	0.2894

Framework versions

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