spcl / daceml

A Data-Centric Compiler for Machine Learning
https://daceml.readthedocs.io
BSD 3-Clause "New" or "Revised" License
81 stars 14 forks source link

Change names of outputs to __return_{i}. #109

Closed orausch closed 2 years ago

orausch commented 2 years ago

This is preparatory work for parsing full training steps.

codecov[bot] commented 2 years ago

Codecov Report

Merging #109 (7e2c2f7) into master (3b5ae4d) will increase coverage by 0.09%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #109      +/-   ##
==========================================
+ Coverage   68.05%   68.13%   +0.09%     
==========================================
  Files          59       59              
  Lines        6810     6819       +9     
==========================================
+ Hits         4634     4646      +12     
+ Misses       2176     2173       -3     
Impacted Files Coverage Δ
daceml/autodiff/backward_pass_generator.py 92.42% <ø> (ø)
daceml/onnx/onnx_importer.py 95.40% <100.00%> (+0.28%) :arrow_up:
daceml/torch/dispatchers/common.py 100.00% <100.00%> (ø)
daceml/torch/dispatchers/cpp_torch_extension.py 94.06% <100.00%> (ø)
daceml/torch/dispatchers/ctypes_module.py 100.00% <100.00%> (+3.03%) :arrow_up:

Help us with your feedback. Take ten seconds to tell us how you rate us.