Skip to content

Error verifying Inductor: torch._inductor.config.triton.autotune does not exist #8877

Description

@necropower

Description
Error verifying Inductor: torch._inductor.config.triton.autotune does not exist

Triton Information
What version of Triton are you using?
Triton: 3.7.1

Are you using the Triton container or did you build it yourself?
pre-compiled wheel

To Reproduce
Steps to reproduce the behavior.
install neoforge on stablematrix, be on this configuration:
Python: 3.12.12 (main, Feb 3 2026, 22:51:04) [Clang 21.1.4 ]
PyTorch: 2.12.1+cu130
CUDA : True
Version CUDA: 13.0
Backend cuDNN: 92000
GPU: NVIDIA GeForce RTX 5090,
install sageattn3, try to generate a image with 315 tokens on positive and 47 on negative prompts, and enable the max-autotune-no-cudagraphs and the following flags:
Launching Web UI with arguments: --pin-shared-memory --cuda-malloc --cuda-stream --skip-python-version-check --sage-function sageattn3 --expandable-segments --mmap-torch-files --gpu-only

Describe the models (framework, inputs, outputs), ideally include the model configuration file (if using an ensemble include the model configuration file for that as well).
waillustirousSDXL_v90.safetensors

Expected behavior
A clear and concise description of what you expected to happen.
should be able to compile the torch compile integrated and resolve the next prompts with speed.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions