xlm-roberta-base-finetuned-panx-hi

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

• Loss: 0.5351
• F1: 0.8699

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: 24
• eval_batch_size: 24
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1
0.843	1.0	86	0.3921	0.7206
0.3751	2.0	172	0.3292	0.7381
0.2437	3.0	258	0.3049	0.8255
0.1963	4.0	344	0.2865	0.8255
0.1519	5.0	430	0.3165	0.8422
0.1332	6.0	516	0.3019	0.8433
0.0748	7.0	602	0.3290	0.8474
0.0703	8.0	688	0.3825	0.8517
0.0611	9.0	774	0.4575	0.8518
0.0484	10.0	860	0.4409	0.8481
0.0345	11.0	946	0.4221	0.8597
0.0242	12.0	1032	0.4673	0.8682
0.0178	13.0	1118	0.4826	0.8671
0.0129	14.0	1204	0.5174	0.8443
0.0139	15.0	1290	0.5028	0.8608
0.0063	16.0	1376	0.5224	0.8648
0.0027	17.0	1462	0.5260	0.8682
0.0036	18.0	1548	0.5301	0.8702
0.0028	19.0	1634	0.5372	0.8741
0.0038	20.0	1720	0.5351	0.8699

Framework versions

• Transformers 4.33.1
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3