e3_lr2e-05

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.6436

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: 16
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.9961	0.1404	100	1.9416
2.0472	0.2808	200	1.8589
1.9766	0.4212	300	1.8095
1.9319	0.5616	400	1.7736
1.897	0.7021	500	1.7447
1.8743	0.8425	600	1.7370
1.86	0.9829	700	1.7156
1.8431	1.1233	800	1.7071
1.8217	1.2637	900	1.6939
1.8212	1.4041	1000	1.6900
1.8053	1.5445	1100	1.6774
1.7899	1.6849	1200	1.6736
1.799	1.8254	1300	1.6644
1.7845	1.9658	1400	1.6559
1.7704	2.1062	1500	1.6531
1.776	2.2466	1600	1.6528
1.773	2.3870	1700	1.6417
1.7632	2.5274	1800	1.6452
1.7451	2.6678	1900	1.6460
1.7505	2.8088	2000	1.6455
1.7602	2.9492	2100	1.6399

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1