furina_seed42_eng_kin_amh

This model is a fine-tuned version of yihongLiu/furina on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0191
• Spearman Corr: 0.7822

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	0.65	200	0.0386	0.5856
No log	1.3	400	0.0256	0.6781
No log	1.95	600	0.0236	0.7129
0.0506	2.61	800	0.0225	0.7250
0.0506	3.26	1000	0.0225	0.7343
0.0506	3.91	1200	0.0215	0.7338
0.022	4.56	1400	0.0210	0.7492
0.022	5.21	1600	0.0219	0.7512
0.022	5.86	1800	0.0199	0.7600
0.016	6.51	2000	0.0209	0.7623
0.016	7.17	2200	0.0201	0.7579
0.016	7.82	2400	0.0203	0.7589
0.0115	8.47	2600	0.0204	0.7604
0.0115	9.12	2800	0.0201	0.7655
0.0115	9.77	3000	0.0201	0.7647
0.0089	10.42	3200	0.0195	0.7652
0.0089	11.07	3400	0.0191	0.7728
0.0089	11.73	3600	0.0197	0.7685
0.0072	12.38	3800	0.0190	0.7709
0.0072	13.03	4000	0.0192	0.7707
0.0072	13.68	4200	0.0189	0.7734
0.006	14.33	4400	0.0193	0.7743
0.006	14.98	4600	0.0191	0.7756
0.006	15.64	4800	0.0189	0.7757
0.0051	16.29	5000	0.0193	0.7733
0.0051	16.94	5200	0.0189	0.7796
0.0051	17.59	5400	0.0185	0.7810
0.0046	18.24	5600	0.0194	0.7781
0.0046	18.89	5800	0.0191	0.7787
0.0046	19.54	6000	0.0186	0.7789
0.0041	20.2	6200	0.0188	0.7798
0.0041	20.85	6400	0.0189	0.7796
0.0041	21.5	6600	0.0187	0.7797
0.0037	22.15	6800	0.0189	0.7787
0.0037	22.8	7000	0.0191	0.7822

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1