mirror of https://github.com/llvm/torch-mlir
6524838bcb
Previously, it could only handle the situations where outputsize == (1, 1) or outputsize == (input_H, input_W). Now it supports all situations where input_H % output_H== 0 && input_W % output_W == 0 |
||
---|---|---|
.. | ||
CAPI | ||
Conversion | ||
Dialect | ||
RefBackend | ||
python | ||
CMakeLists.txt | ||
lit.cfg.py | ||
lit.site.cfg.py.in |