xlmr_experiment

This model is a fine-tuned version of xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5843
• Accuracy: 0.8961
• F1: 0.9086
• Precision: 0.9485
• Recall: 0.872

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
5.6500	0.6319	100	1.2349	0.7282	0.7958	0.7171	0.894
4.4099	1.2591	200	1.0472	0.7598	0.7743	0.8735	0.6953
3.6862	1.8910	300	0.7443	0.8471	0.8738	0.8551	0.8933
2.7468	2.5182	400	0.5972	0.8819	0.8990	0.9110	0.8873
2.3632	3.1453	500	0.5091	0.9012	0.9160	0.9234	0.9087
1.7778	3.7773	600	0.4864	0.9087	0.9210	0.9453	0.898
1.3475	4.4044	700	0.4459	0.9202	0.9315	0.9482	0.9153

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.3
• Tokenizers 0.22.2