Skip to content

Assertion srcIndex < srcSelectDimSize failed in the inference.py #141

@step-out

Description

@step-out

When I run the inference.py on the hundreds of audio inputs with the A1A2 task, I meet the following question:

... (correct audio output)

../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [16,0,0], thread: [88,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [16,0,0], thread: [89,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [16,0,0], thread: [90,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [16,0,0], thread: [91,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [16,0,0], thread: [92,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [16,0,0], thread: [93,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [16,0,0], thread: [94,0,0] Assertion `srcIndex < srcSelectDimSize` failed.
../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [16,0,0], thread: [95,0,0] Assertion `srcIndex < srcSelectDimSize` failed.

[error] failed to process /data/21.mp3
[error] failed to process /data/22.mp3
[error] failed to process /data/23.mp3
[error] failed to process /data/24.mp3
[error] failed to process /data/25.mp3
[error] failed to process /data/26.mp3
...

I meet the situation on two different machines; the machines' information is:

System 1:
Distributor ID: Ubuntu
Description:    Ubuntu 22.04.4 LTS
Release:        22.04
Codename:       jammy

GPU 1:
NVIDIA H20
System 2:
CENTOS_MANTISBT_PROJECT="CentOS-8"
CENTOS_MANTISBT_PROJECT_VERSION="8"
REDHAT_SUPPORT_PRODUCT="centos"
REDHAT_SUPPORT_PRODUCT_VERSION="8"
NAME_ORIG="CentOS Linux"

GPU 2:
Tesla T4

Has anyone encountered this problem?

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