Edit model card

xlm-roberta-base_hau_corr_5e-06

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0237
  • Spearman Corr: 0.7615

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 0.95 200 0.0254 0.7574
No log 1.91 400 0.0249 0.7595
0.0025 2.86 600 0.0243 0.7626
0.0025 3.82 800 0.0251 0.7601
0.0023 4.77 1000 0.0241 0.7589
0.0023 5.73 1200 0.0236 0.7597
0.0023 6.68 1400 0.0243 0.7610
0.0023 7.64 1600 0.0251 0.7637
0.0022 8.59 1800 0.0251 0.7599
0.0022 9.55 2000 0.0255 0.7603
0.0021 10.5 2200 0.0252 0.7606
0.0021 11.46 2400 0.0244 0.7628
0.002 12.41 2600 0.0257 0.7644
0.002 13.37 2800 0.0250 0.7610
0.002 14.32 3000 0.0254 0.7600
0.002 15.27 3200 0.0250 0.7618
0.0019 16.23 3400 0.0249 0.7617
0.0019 17.18 3600 0.0247 0.7602
0.0019 18.14 3800 0.0245 0.7624
0.0019 19.09 4000 0.0244 0.7639
0.0018 20.05 4200 0.0237 0.7615

Framework versions

  • Transformers 4.37.2
  • Pytorch 2.2.0+cu121
  • Datasets 2.17.0
  • Tokenizers 0.15.2
Downloads last month
0
Safetensors
Model size
278M params
Tensor type
F32
·

Finetuned from