xlm-roberta-base_latin_original_kin-amh-eng_train_spearman_corr

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

• Loss: 0.0296
• Spearman Corr: 0.7280

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: 128
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Spearman Corr
No log	1.75	200	0.0262	0.6814
0.0524	3.51	400	0.0244	0.7063
0.0217	5.26	600	0.0268	0.7190
0.0158	7.02	800	0.0277	0.7417
0.0124	8.77	1000	0.0271	0.7239
0.0098	10.53	1200	0.0255	0.7280
0.0081	12.28	1400	0.0296	0.7328
0.0068	14.04	1600	0.0293	0.7404
0.0068	15.79	1800	0.0269	0.7349
0.006	17.54	2000	0.0300	0.7253
0.005	19.3	2200	0.0266	0.7324
0.0047	21.05	2400	0.0296	0.7280

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.2