Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

how to resolve AttributeError: 'Config' object has no attribute 'all_kwargs'? #13

Open
KimmiShi opened this issue Dec 12, 2024 · 1 comment

Comments

@KimmiShi
Copy link

KimmiShi commented Dec 12, 2024

Hi, I am trying to run ml-cross-entropy/benchmark, but I encountered the following error:

Traceback (most recent call last):
  File "/work/ml-cross-entropy/benchmark/__main__.py", line 15, in <module>
    from cut_cross_entropy import linear_cross_entropy
  File "/work/ml-cross-entropy/cut_cross_entropy/__init__.py", line 2, in <module>
    from cut_cross_entropy.linear_cross_entropy import (
  File "/work/ml-cross-entropy/cut_cross_entropy/linear_cross_entropy.py", line 20, in <module>
    from cut_cross_entropy.cce import cce_linear_cross_entropy
  File "/work/ml-cross-entropy/cut_cross_entropy/cce.py", line 7, in <module>
    from cut_cross_entropy.cce_backward import cce_backward_kernel
  File "/work/ml-cross-entropy/cut_cross_entropy/cce_backward.py", line 237, in <module>
    _cce_backward_kernel = cce_backward_autotune()(_cce_backward_kernel)  # type: ignore
  File "/work/ml-cross-entropy/cut_cross_entropy/tl_autotune.py", line 422, in cce_backward_autotune
    return _heuristics_from_config(_cce_backward_best_config())
  File "/work/ml-cross-entropy/cut_cross_entropy/tl_autotune.py", line 368, in _heuristics_from_config
    return triton.heuristics({k: (lambda args, _v=v: _v) for k, v in config.all_kwargs().items()})
AttributeError: 'Config' object has no attribute 'all_kwargs'

my env info:

Python 3.10.4
torch: 2.4.0+cu118
triton: 3.0.0
GPU: A800
@KimmiShi
Copy link
Author

triton version should be higher than 3.0.0, I don't know the minimal version that could work, I tried 3.1.0 and it works.

perhaps the triton version requirement in readme needs update.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant