furina_esp_loss_5e-06

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

• Loss: 0.0212
• Spearman Corr: 0.7689

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	0.94	200	0.0221	0.7657
No log	1.89	400	0.0230	0.7642
0.0024	2.83	600	0.0224	0.7652
0.0024	3.77	800	0.0216	0.7663
0.0023	4.72	1000	0.0223	0.7644
0.0023	5.66	1200	0.0218	0.7638
0.0023	6.6	1400	0.0229	0.7641
0.0023	7.55	1600	0.0219	0.7640
0.0022	8.49	1800	0.0230	0.7634
0.0022	9.43	2000	0.0228	0.7634
0.0022	10.38	2200	0.0224	0.7661
0.0022	11.32	2400	0.0211	0.7680
0.0028	12.26	2600	0.0216	0.7687
0.0028	13.21	2800	0.0219	0.7669
0.0028	14.15	3000	0.0218	0.7697
0.0028	15.09	3200	0.0216	0.7665
0.0027	16.04	3400	0.0215	0.7664
0.0027	16.98	3600	0.0219	0.7683
0.0027	17.92	3800	0.0215	0.7691
0.0026	18.87	4000	0.0212	0.7689

Framework versions

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