Hugging Face's logo

xshubhamx
/
bart-large-lora

TensorBoard
Safetensors
Generated from Trainer
Model card Files Metrics Community
bart-large-lora / README.md
xshubhamx's picture
xshubhamx
Upload README.md with huggingface_hub
66b1e21 verified
preview code
|
raw
history blame contribute delete
No virus
6.07 kB
metadata
license: apache-2.0
tags:
  - generated_from_trainer
base_model: facebook/bart-large
metrics:
  - accuracy
  - precision
  - recall
model-index:
  - name: bart-base-lora
    results: []

bart-base-lora

This model is a fine-tuned version of facebook/bart-large on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6884
  • Accuracy: 0.8172
  • Precision: 0.8132
  • Recall: 0.8172
  • Precision Macro: 0.7584
  • Recall Macro: 0.7412
  • Macro Fpr: 0.0164
  • Weighted Fpr: 0.0157
  • Weighted Specificity: 0.9755
  • Macro Specificity: 0.9862
  • Weighted Sensitivity: 0.8172
  • Macro Sensitivity: 0.7412
  • F1 Micro: 0.8172
  • F1 Macro: 0.7417
  • F1 Weighted: 0.8124

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: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 15

Training results

Training Loss Epoch Step Validation Loss Accuracy Precision Recall Precision Macro Recall Macro Macro Fpr Weighted Fpr Weighted Specificity Macro Specificity Weighted Sensitivity Macro Sensitivity F1 Micro F1 Macro F1 Weighted
No log 1.0 160 0.9525 0.7157 0.6788 0.7157 0.3875 0.4416 0.0285 0.0276 0.9642 0.9787 0.7157 0.4416 0.7157 0.3958 0.6835
No log 2.0 321 0.7733 0.7413 0.7296 0.7413 0.4491 0.4687 0.0252 0.0243 0.9668 0.9805 0.7413 0.4687 0.7413 0.4337 0.7231
No log 3.0 482 0.7105 0.7738 0.7631 0.7738 0.5565 0.5408 0.0212 0.0205 0.9725 0.9831 0.7738 0.5408 0.7738 0.5271 0.7611
1.08 4.0 643 0.7539 0.7576 0.7584 0.7576 0.5791 0.5613 0.0234 0.0223 0.9681 0.9817 0.7576 0.5613 0.7576 0.5497 0.7438
1.08 5.0 803 0.6978 0.7831 0.7900 0.7831 0.7410 0.6492 0.0203 0.0194 0.9710 0.9836 0.7831 0.6492 0.7831 0.6354 0.7703
1.08 6.0 964 0.5920 0.8156 0.8053 0.8156 0.7051 0.6889 0.0166 0.0159 0.9746 0.9860 0.8156 0.6889 0.8156 0.6860 0.8088
0.5581 7.0 1125 0.6231 0.8187 0.8178 0.8187 0.7627 0.7425 0.0162 0.0156 0.9766 0.9864 0.8187 0.7425 0.8187 0.7393 0.8147
0.5581 8.0 1286 0.6291 0.8141 0.8134 0.8141 0.7636 0.7307 0.0167 0.0160 0.9758 0.9860 0.8141 0.7307 0.8141 0.7329 0.8089
0.5581 9.0 1446 0.6226 0.8242 0.8212 0.8242 0.7666 0.7340 0.0158 0.0150 0.9760 0.9867 0.8242 0.7340 0.8242 0.7365 0.8191
0.3924 10.0 1607 0.6728 0.8110 0.8123 0.8110 0.7418 0.7289 0.0170 0.0164 0.9762 0.9858 0.8110 0.7289 0.8110 0.7240 0.8048
0.3924 11.0 1768 0.6805 0.8095 0.8123 0.8095 0.7390 0.7303 0.0173 0.0165 0.9752 0.9856 0.8095 0.7303 0.8095 0.7263 0.8026
0.3924 12.0 1929 0.6710 0.8133 0.8137 0.8133 0.7396 0.7306 0.0168 0.0161 0.9759 0.9859 0.8133 0.7306 0.8133 0.7284 0.8090
0.2929 13.0 2089 0.6740 0.8187 0.8170 0.8187 0.7644 0.7360 0.0162 0.0156 0.9761 0.9863 0.8187 0.7360 0.8187 0.7368 0.8151
0.2929 14.0 2250 0.6823 0.8180 0.8159 0.8180 0.7657 0.7336 0.0164 0.0156 0.9753 0.9862 0.8180 0.7336 0.8180 0.7361 0.8137
0.2929 14.93 2400 0.6884 0.8172 0.8132 0.8172 0.7584 0.7412 0.0164 0.0157 0.9755 0.9862 0.8172 0.7412 0.8172 0.7417 0.8124

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.19.0
  • Tokenizers 0.15.1