5b4deff7a40ccd4d93195d9c7f1ba494

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

• Loss: 0.6827
• Data Size: 0.25
• Epoch Runtime: 31.8932
• 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	1.2886	0	7.2508	0.0601	0.0378
No log	1	619	0.8110	0.0078	8.0386	0.7672	0.2894
No log	2	1238	0.6233	0.0156	9.2341	0.7672	0.2894
0.0167	3	1857	0.6897	0.0312	11.2092	0.7672	0.2894
0.0167	4	2476	0.7081	0.0625	14.8542	0.7672	0.2894
0.6771	5	3095	0.6825	0.125	20.1206	0.7672	0.2894
0.0643	6	3714	0.6827	0.25	31.8932	0.7672	0.2894

Framework versions

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