Skip to content

hack7

Javier Duarte requested to merge github/fork/vloncar/hack7 into master

Created by: vloncar

This PR is the culmination of several months of development and has many changes. The ones I remember:

  • tf_to_hls tool for converting tensorflow models
  • Experimental support for larger Conv1D/2D layers
  • QKeras support (limited to binary/ternary for now)
  • API enhancements (custom layers, multiple backends)
  • Profiling support
  • hls4ml report command
  • Support for all-in-one h5 files (obtained with Keras save() function, without the need for separate json)
  • many more smaller changes I forgot.

Feedback is welcome

Merge request reports

Loading