furina_seed42_eng_amh_esp_basic_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: 0.0213
• Spearman Corr: 0.7510

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	1.59	200	0.0439	0.1550
0.1335	3.17	400	0.0317	0.6373
0.0485	4.76	600	0.0210	0.6957
0.0295	6.35	800	0.0249	0.7234
0.0295	7.94	1000	0.0196	0.7301
0.0243	9.52	1200	0.0195	0.7405
0.0221	11.11	1400	0.0214	0.7448
0.02	12.7	1600	0.0237	0.7426
0.0187	14.29	1800	0.0198	0.7470
0.0187	15.87	2000	0.0208	0.7501
0.0176	17.46	2200	0.0240	0.7495
0.0163	19.05	2400	0.0212	0.7518
0.0158	20.63	2600	0.0237	0.7475
0.0154	22.22	2800	0.0213	0.7510

Framework versions

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