Conversation
|
@mxnet-label-bot add [CI] |
|
Hey @hzfan thanks for the PR. Have we tested on CI-dev if it passes the nightly? That's generally a good indicator. I can do it for you if you haven't. |
python/mxnet/tvmop.py
Outdated
| try: | ||
| _CONF_TVM_OP = find_conf_path("tvmop") | ||
| except RuntimeError as e: | ||
| logging.warning(e) |
There was a problem hiding this comment.
Could you add comments on what's the implication of this? Why we just print a warning. It's not very clear when going through the code.
There was a problem hiding this comment.
Comments added. The config file records available schedules, but besides these, a default schedule is also available, so when the conf is missing, we can fallback to default.
There was a problem hiding this comment.
Could you still add a message that explains this to the user? otherwise, they just see an error as warning without context.
Hi @ChaiBapchya , haven't tested on CI-dev nightly yet. Appreciate it if if you could help. |
|
Oups I tried to suggest a change, but instead directly committed to your branch. Sorry about that! I still have to figure out how the proposal systems works <.< |
Sure. I have pointed the NightlyTestForBinaries to test out for conf_pr branch in your repository. It passed. |
There was a problem hiding this comment.
Thanks for the fix, now that the nightly passed (thanks @ChaiBapchya) lets merge this PR @reminisce @ptrendx
* Remove dependency on tvmop.conf * Fix binaries dependencies for ni nightly * Add comments * Update tvmop.py * Fix rebase
…, #16792) (#16832) * Fix nightly build (#16773) * Remove dependency on tvmop.conf * Fix binaries dependencies for ni nightly * Add comments * Update tvmop.py * Fix rebase * Fix (#16781) * Speed fused_op compilation by caching ptx and jit-compiled functions (#16783) * [Numpy] Fix collect_params().zero_grad() in gluon numpy interface (#16716) * fix zero_grad * Update parameter.py * add test * fix * Mixed data type binary ops (#16699) * support mixed-precision binary operations * improvement for documentations and error messages * Support boolean elemwise/broadcast binary add, multiply and true_divide (#16728) * support pure boolean elemwise/broadcast binary op * switch to unique_tpr * fix the test error * Fix rtrue_divide grad (#16769) * Fix rtrue_divide_scalar * More tests * Fix numpy-compatible mean output type for integer inputs (#16792) * fix mean output type for integer inputs * enable for windows
Description
Include
tvmop.confas binary dependencies.Checklist
Essentials
Please feel free to remove inapplicable items for your PR.
Changes
Comments
Thank @yzhliu for guidance and review.