xnli_xlm_r_only_tr

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

• Loss: 0.8306
• Accuracy: 0.7498

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: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.769	1.0	3068	0.6296	0.7281
0.6402	2.0	6136	0.5829	0.7586
0.579	3.0	9204	0.6268	0.7474
0.5258	4.0	12272	0.6304	0.7478
0.4796	5.0	15340	0.6619	0.7466
0.4363	6.0	18408	0.7173	0.7438
0.398	7.0	21476	0.7551	0.7498
0.3666	8.0	24544	0.7922	0.7478
0.3403	9.0	27612	0.8081	0.7534
0.3216	10.0	30680	0.8306	0.7498

Framework versions

• Transformers 4.24.0
• Pytorch 1.13.0
• Datasets 2.6.1
• Tokenizers 0.13.1