furina_arb_corr_2e-05

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

• Loss: 0.0221
• Spearman Corr: 0.7735

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	0.85	200	0.0240	0.7691
No log	1.69	400	0.0235	0.7701
0.0027	2.54	600	0.0237	0.7705
0.0027	3.38	800	0.0235	0.7689
0.0026	4.23	1000	0.0224	0.7702
0.0026	5.07	1200	0.0233	0.7759
0.0026	5.92	1400	0.0230	0.7711
0.0023	6.77	1600	0.0236	0.7702
0.0023	7.61	1800	0.0208	0.7744
0.0021	8.46	2000	0.0223	0.7733
0.0021	9.3	2200	0.0225	0.7775
0.002	10.15	2400	0.0213	0.7746
0.002	10.99	2600	0.0219	0.7763
0.002	11.84	2800	0.0218	0.7750
0.0021	12.68	3000	0.0216	0.7739
0.0021	13.53	3200	0.0221	0.7732
0.0019	14.38	3400	0.0218	0.7759
0.0019	15.22	3600	0.0210	0.7735
0.0018	16.07	3800	0.0221	0.7735

Framework versions

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