furina_seed42_eng_kin_amh_basic_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: 0.0207
• Spearman Corr: 0.7614

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	0.0250	0.6539
0.0825	3.51	400	0.0215	0.7101
0.0225	5.26	600	0.0232	0.7326
0.0158	7.02	800	0.0218	0.7517
0.0126	8.77	1000	0.0218	0.7565
0.0104	10.53	1200	0.0205	0.7582
0.0088	12.28	1400	0.0200	0.7707
0.0074	14.04	1600	0.0213	0.7565
0.0074	15.79	1800	0.0202	0.7601
0.0065	17.54	2000	0.0221	0.7605
0.0059	19.3	2200	0.0221	0.7588
0.0055	21.05	2400	0.0214	0.7561
0.005	22.81	2600	0.0219	0.7574
0.0047	24.56	2800	0.0208	0.7606
0.0045	26.32	3000	0.0207	0.7614

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.2