furina_original_kin-hau-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: 0.0267
• Spearman Corr: 0.7425

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	1.59	200	0.0442	0.5364
0.0796	3.19	400	0.0280	0.6912
0.0275	4.78	600	0.0329	0.7307
0.0199	6.37	800	0.0256	0.7473
0.015	7.97	1000	0.0275	0.7502
0.015	9.56	1200	0.0265	0.7396
0.0127	11.16	1400	0.0264	0.7484
0.0104	12.75	1600	0.0255	0.7467
0.0089	14.34	1800	0.0256	0.7512
0.0079	15.94	2000	0.0270	0.7457
0.0079	17.53	2200	0.0264	0.7420
0.0069	19.12	2400	0.0276	0.7444
0.0062	20.72	2600	0.0265	0.7383
0.0059	22.31	2800	0.0264	0.7416
0.0055	23.9	3000	0.0269	0.7439
0.0055	25.5	3200	0.0268	0.7415
0.0051	27.09	3400	0.0267	0.7425

Framework versions

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