Another idea @zxybazh and I have been discussing is the possibility to unify
Python and CXX logging system via packed functions, i.e. the CXX logging system
could potentially call back to python's logging module
---
[Visit
Topic](https://discuss.tvm.apache.org/t/rfc-better-tvm-logger-in-c
I like this proposal a lot. Here are my two cents in terms of use cases.
In the C++ side, PyTorch has two logging systems:
- `LOG(LEVEL)`: This is what we familiar with, including DEBUG, INFO, WARNING,
and FATAL.
- `VLOG(INT_LEVEL)`: This is only for debugging, `INT_LEVEL` could range from 0
t
This is definitely extremely important! I would love to further discuss with
you on the environment variables used here. Do you think it’s possible to unify
DMLC_ and TVM_ environment variables? Is it possible to minimize the number of
env variables to use? Is there any precedent we could refe
I think this is a great initiative @pzq. The only suggestion I'd put on top of
it, is that we also can make the Python side to - generally - respect the same
rules, environment variables, etc. Just to keep in mind that we can offer a
better experience for our users if these rules apply more wi
For now, tvm uses the dmlc log in `dmlc-core` by default, which lacks good log
level management. Below is some underperformance in tvm.
1. `LOG(INFO)` shares the same severity with `LOG(WARNING)` and `LOG(ERROR)`,
2. `LOG(DEBUG)` does not exist. Instead, tvm contains a macro `TVM_LOG_DEBUG`
an
