Finetune Versions Save

Scikit-learn style model finetuning for NLP

0.10.0

7 months ago

0.8.6

4 years ago

0.8.5

4 years ago

0.7.2

4 years ago
  • Auxiliary Information
  • Chunk long sequences for models other than SequenceLabeler
  • Bug fix release

0.7.0

4 years ago
  • GPT2-medium support
  • Support for adapter finetuning strategy
  • Add DeploymentModel class

0.6.4

4 years ago

2019 v0.6.4 Tue May 14th

  • Add BERT support
  • Fix: memory issue
  • Support basic attention visualization

0.6.3

5 years ago

v0.6.3 Thu May 2nd

  • Option to control maximum percentage of memory used
  • Attention weight visualization for GPT
  • Assorted bug fixes

0.6.2

5 years ago

v0.6.2 Thu Apr 4th

  • Cached predict fix
  • Model download fix

0.6.0

5 years ago

v0.6.0 Thu Mar 28th

  • Support for new base models
  • GPT2 support
  • Multi-task learning support
  • Word CNN Baseline
  • Code base refactor
  • CUDA 10

0.5.14

5 years ago