variant-tapt_freeze_llrd_ww_mask-LR_5e-05

This model is a fine-tuned version of microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext on the Mardiyyah/TAPT_Variant_FT dataset. It achieves the following results on the evaluation set:

  • Loss: 1.4054
  • Accuracy: 0.7228

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-05
  • train_batch_size: 16
  • eval_batch_size: 16
  • seed: 3407
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 32
  • optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.06
  • num_epochs: 20
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.6473 1.0 19 1.6487 0.6997
1.6283 2.0 38 1.6091 0.7057
1.5706 3.0 57 1.5814 0.7040
1.5255 4.0 76 1.5358 0.7120
1.5047 5.0 95 1.5496 0.7085
1.4924 6.0 114 1.4409 0.7167
1.463 7.0 133 1.4329 0.7179
1.4172 8.0 152 1.5335 0.7107
1.4397 9.0 171 1.4242 0.7189
1.4296 10.0 190 1.4802 0.7081
1.399 11.0 209 1.4528 0.7215
1.4182 12.0 228 1.4391 0.7149
1.4079 13.0 247 1.4564 0.7185
1.4032 14.0 266 1.4057 0.7202
1.4094 15.0 285 1.4528 0.7146
1.3706 16.0 304 1.4888 0.7160
1.3522 17.0 323 1.4005 0.7207
1.3732 18.0 342 1.4291 0.7118
1.3696 19.0 361 1.4123 0.7221
1.3729 20.0 380 1.4188 0.7199

Framework versions

  • Transformers 4.48.2
  • Pytorch 2.4.1+cu121
  • Datasets 3.0.2
  • Tokenizers 0.21.0
Downloads last month
1
Safetensors
Model size
0.1B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Mardiyyah/variant-tapt_freeze_llrd_ww_mask-LR_5e-05