Metrics

  • loss: 0.6616
  • accuracy: 0.8443
  • precision: 0.8473
  • recall: 0.8443
  • precision_macro: 0.8200
  • recall_macro: 0.7913
  • macro_fpr: 0.0134
  • weighted_fpr: 0.0130
  • weighted_specificity: 0.9802
  • macro_specificity: 0.9883
  • weighted_sensitivity: 0.8443
  • macro_sensitivity: 0.7913
  • f1_micro: 0.8443
  • f1_macro: 0.7980
  • f1_weighted: 0.8435
  • runtime: 28.7426
  • samples_per_second: 44.9160
  • steps_per_second: 5.6360

legal-InLegal-merge-ties

legal-InLegal-merge-ties is a merge of the following models using mergekit:

🧩 Configuration

models:
  - model: xshubhamx/legal-bert-base-uncased
    parameters:
      density: 0.5
      weight: 0.5
  - model: xshubhamx/InLegalBERT
    parameters:
      density: 0.5
      weight: 0.5

merge_method: ties
base_model: xshubhamx/legal-bert-base-uncased
parameters:
  normalize: false
  int8_mask: true
dtype: float16
Downloads last month
107
Safetensors
Model size
109M params
Tensor type
FP16
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.