Skip to content
master
Switch branches/tags
Go to file
Code

Latest commit

* [FIX] Vivado HLS 2019.2 crashes while synthesizing
Issue: Array subscript could not be verified to always be non-negative

* Moving conifer.backends.vivadohls to xilinxhls
Generalize package for vivadohls\vitishls interoperability

* Backward compatibility
Aliases the old conifer.backends.vivadohls
Adds the corresponding conifer.backends.vitishls

* Update defaults to backends.xilinxhls
With automatic detection of Vivado\Vitis HLS tool in path.
In case of both, Vivado is preferred.

* Add pipeline option

Co-authored-by: Andrea Damiani <1655566+andredami@users.noreply.github.com>
e1e5827

Git stats

Files

Permalink
Failed to load latest commit information.

README.md

conifer

Conifer translates trained Boosted Decision Trees to FPGA firmware for extreme low latency inference. Check the examples directory for examples to get started with.

Currently models from sklearn, xgboost, and TMVA are supported. FPGA firmware can be produced in Xilinx Vivado HLS or VHDL.

See our paper: https://arxiv.org/abs/2002.02534

Conifer originated as a development for https://hls-fpga-machine-learning.github.io/hls4ml/, and borrows heavily from the code and ideas developed for it.

Installation

git clone https://github.com/thesps/conifer.git
cd conifer
pip install .

About

Fast inference of Boosted Decision Trees in FPGAs

Topics

Resources

License

Releases

No releases published

Packages

No packages published