Edit model card

furina_latin_kin-amh-eng_train_spearman_corr

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

  • Loss: 0.0287
  • Spearman Corr: 0.7455

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 0.59 200 0.0534 0.5665
No log 1.17 400 0.0303 0.6812
No log 1.76 600 0.0374 0.6929
0.0479 2.35 800 0.0244 0.7282
0.0479 2.93 1000 0.0333 0.7172
0.0479 3.52 1200 0.0287 0.7167
0.0233 4.11 1400 0.0287 0.7330
0.0233 4.69 1600 0.0297 0.7176
0.0233 5.28 1800 0.0255 0.7429
0.0233 5.87 2000 0.0320 0.7385
0.0165 6.45 2200 0.0273 0.7325
0.0165 7.04 2400 0.0262 0.7489
0.0165 7.62 2600 0.0343 0.7388
0.0121 8.21 2800 0.0258 0.7398
0.0121 8.8 3000 0.0298 0.7398
0.0121 9.38 3200 0.0303 0.7370
0.0121 9.97 3400 0.0316 0.7394
0.0095 10.56 3600 0.0295 0.7395
0.0095 11.14 3800 0.0299 0.7399
0.0095 11.73 4000 0.0287 0.7455

Framework versions

  • Transformers 4.37.2
  • Pytorch 2.2.0+cu121
  • Datasets 2.17.0
  • Tokenizers 0.15.2
Downloads last month
3
Safetensors
Model size
394M params
Tensor type
F32
·

Finetuned from