Skip to content

DeepDetect v0.9.6 - BERT Pytorch + TensorRT 7 OSS + new OCR models
Compare
Choose a tag to compare
@beniz beniz released this 05 Oct 11:28
· 990 commits to master since this release
v0.9.6
65fbb36
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.

Features & updates

  • BERT + GPT2 training / finetuning / inference with Torch C++ API #673
  • TensorRT 5, 6 & 7 OSS #683
  • lr dropout #668 as an implementation of https://arxiv.org/abs/1912.00144 (also see jolibrain/caffe#58)
  • confidence on OCR/CTC output prediction #694
  • support for logger inside chain actions #693
  • added support for OCR models (CRNN) Squeeze-Excitation ResNet-50 and ResNeXt-50 #685

API changes

  • mllib.lr_dropout#668
  • image rotation action in chains #677

Bug fixes

  • inverted ymin and ymax in detection model output, #690
  • chain action crop failing when bbox is not within the image #693
  • fixed bug in raw measure data output for model calibration #679
  • fixed error message on wrong TensorRT model filename #681
  • fixed potentially empty chain output #692
Assets 2
Loading