xlm-roberta-base_lr5e-06_seed42_basic_original_amh-esp-eng_train

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: 0.0208
• Spearman Corr: 0.7694

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-06
• 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	1.59	200	0.0178	0.7278
0.0931	3.17	400	0.0197	0.7452
0.0276	4.76	600	0.0183	0.7513
0.0243	6.35	800	0.0158	0.7667
0.0243	7.94	1000	0.0182	0.7692
0.021	9.52	1200	0.0189	0.7704
0.0192	11.11	1400	0.0184	0.7656
0.0176	12.7	1600	0.0226	0.7691
0.0163	14.29	1800	0.0225	0.7696
0.0163	15.87	2000	0.0207	0.7718
0.0156	17.46	2200	0.0242	0.7697
0.0143	19.05	2400	0.0207	0.7664
0.0138	20.63	2600	0.0201	0.7721
0.013	22.22	2800	0.0208	0.7694

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.2