furina_seed42_eng_esp_kin_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: 0.0246
• Spearman Corr: 0.7089

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	1.63	200	0.0555	0.1927
0.1485	3.27	400	0.0295	0.5212
0.0499	4.9	600	0.0270	0.5827
0.0297	6.53	800	0.0263	0.6144
0.025	8.16	1000	0.0274	0.6415
0.025	9.8	1200	0.0245	0.6578
0.0229	11.43	1400	0.0271	0.6716
0.0209	13.06	1600	0.0256	0.6775
0.0194	14.69	1800	0.0237	0.6853
0.0183	16.33	2000	0.0245	0.6979
0.0183	17.96	2200	0.0260	0.6951
0.0172	19.59	2400	0.0256	0.6999
0.0165	21.22	2600	0.0240	0.7056
0.0158	22.86	2800	0.0239	0.7050
0.0154	24.49	3000	0.0244	0.7078
0.0148	26.12	3200	0.0246	0.7087
0.0148	27.76	3400	0.0246	0.7089

Framework versions

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