furina_seed42_eng_amh_hau_cross_0.0001

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

• Loss: 0.0312
• Spearman Corr: 0.7298

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: 0.0001
• 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	0.52	200	0.0443	0.1783
No log	1.04	400	0.0333	0.5121
No log	1.55	600	0.0424	0.5339
0.0522	2.07	800	0.0398	0.5674
0.0522	2.59	1000	0.0328	0.6002
0.0522	3.11	1200	0.0313	0.6285
0.0522	3.63	1400	0.0292	0.6480
0.0361	4.15	1600	0.0297	0.6471
0.0361	4.66	1800	0.0298	0.6724
0.0361	5.18	2000	0.0308	0.7280
0.0361	5.7	2200	0.0262	0.7299
0.0258	6.22	2400	0.0255	0.7406
0.0258	6.74	2600	0.0284	0.7288
0.0258	7.25	2800	0.0295	0.7337
0.0258	7.77	3000	0.0300	0.7393
0.0164	8.29	3200	0.0271	0.7451
0.0164	8.81	3400	0.0319	0.7359
0.0164	9.33	3600	0.0261	0.7314
0.0164	9.84	3800	0.0290	0.7265
0.0105	10.36	4000	0.0312	0.7298

Framework versions

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