furina_latin_original_kin-amh-eng_train_spearman_corr

This model is a fine-tuned version of yihongLiu/furina on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0314
• Spearman Corr: 0.7391

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	0.0347	0.6054
0.0869	3.51	400	0.0244	0.7064
0.0226	5.26	600	0.0236	0.7265
0.0164	7.02	800	0.0271	0.7412
0.0132	8.77	1000	0.0264	0.7449
0.0107	10.53	1200	0.0300	0.7453
0.0087	12.28	1400	0.0314	0.7445
0.0078	14.04	1600	0.0252	0.7416
0.0078	15.79	1800	0.0276	0.7428
0.0067	17.54	2000	0.0312	0.7400
0.0061	19.3	2200	0.0307	0.7426
0.0057	21.05	2400	0.0318	0.7400
0.0053	22.81	2600	0.0326	0.7363
0.0049	24.56	2800	0.0314	0.7391

Framework versions

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