Skip to content

Latest commit

 

History

History
42 lines (23 loc) · 1022 Bytes

NEWS.md

File metadata and controls

42 lines (23 loc) · 1022 Bytes
Raw

NERDA 1.0.0

  • NERDA model class is now equipped with functions for saving (loading) weights for a fine-tuned NERDA Network to (from) file. See functions model.save_network() and model.load_network_from_file()

NERDA 0.9.7

  • return confidence scores for predictions of all tokens, e.g. model.predict(x, return_confidence=True).

NERDA 0.9.6

  • compute Precision, Recall and Accuracy (optional) with evaluate_performance().
  • improve relative imports inside package.

NERDA 0.9.5

  • ... bugfixes.

NERDA 0.9.4

  • functionality for dynamic quantization, fp32 to fp16, padding parametrized.

NERDA 0.9.2

  • remove precooked DA_BERT_ML_16BIT, include precooked DA_DISTILBERT_ML.

NERDA 0.9.1

  • include 16 bit FP precooked DA_BERT_ML_16BIT.

NERDA 0.9.0

  • Support new versions of transformers (4.x) and torch

NERDA 0.8.7

  • BUGFIX: Restrict torch version.
  • Do not import datasets as part of Precooked Models.
  • Do not load datasets if not provided by user.

NERDA 0.8.6

  • First official release.