Name: Deberta-Finetuning-Results
Creator: Kaggle
License: https://www.apache.org/licenses/

About this Dataset

Deberta-Finetuning-Results

This dataset contains a file summarizing the performances of Deberta on NER task of PIID using different hyper-parameters for both training and testing. The results for each fold are presented as well as the mean performance.
The metrics used are :

Precision
Recall
F1 score (beta=5)

Tables

Deberta Finetuning

@kaggle.jarvisai7_deberta_finetuning_results.deberta_finetuning

16.39 kB
16 rows
20 columns

CREATE TABLE deberta_finetuning (
  "unnamed_0" BIGINT  -- Unnamed: 0,
  "accumulation" BIGINT,
  "batch" BIGINT,
  "epochs" BIGINT,
  "threshold" DOUBLE,
  "effective_batch" BIGINT,
  "f1_fold_0" DOUBLE,
  "f1_fold_1" DOUBLE,
  "f1_fold_2" DOUBLE,
  "f1_fold_3" DOUBLE,
  "mean_f1" DOUBLE,
  "precision_fold_0" DOUBLE,
  "precision_fold_1" DOUBLE,
  "precision_fold_2" DOUBLE,
  "precision_fold_3" DOUBLE,
  "recall_fold_0" DOUBLE,
  "recall_fold_1" DOUBLE,
  "recall_fold_2" DOUBLE,
  "recall_fold_3" DOUBLE,
  "std_f1" DOUBLE
);