[DOCKER][Adreno]Docker infra for Adreno target with CLML support #12833
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@valmat07 can you take a look at the residual-block test case. I see same behaviour with Nvidia (NVIDIA GeForce RTX 3060) and Adreno (Xiaomi Mi 12 - Snapdragon Gen 1) devices. |
New docker for Adreno that inherits gpu docker and amends android-sdk. For CLML, we need to specify ADRENO_OPENCL via environment variable. CLML SDK can be downlaoded from Qualcomm Developer Network by following https://developer.qualcomm.com/blog/accelerate-your-models-our-opencl-ml-sdk Adreno device is shared with host, hence networking is enabled for Adreno Docker.
srkreddy1238
changed the title
driazati
reviewed
Sep 20, 2022
Co-authored-by: driazati <[email protected]>
srkreddy1238
force-pushed
the
main
branch
2 times, most recently
from
5a68065 to
4b1ab12
Compare
valmat07
suggested changes
Sep 22, 2022
There was a problem hiding this comment.
I checked the residual block test on the main tvm line for the Adreno device, it seems after some changes the residual test is not working. Could you also remove skipif:
tvm/tests/python/relay/opencl_texture/test_conv2d_nchw_texture.py
Lines 480 to 483 in 5a68065
elvin-n
suggested changes
Sep 22, 2022
echuraev
approved these changes
Sep 26, 2022
srkreddy1238
requested review from
elvin-n and
valmat07
and removed request for
valmat07 and
elvin-n
srkreddy1238
force-pushed
the
main
branch
2 times, most recently
from
32634f1 to
b7e165e
Compare
masahi
approved these changes
Oct 18, 2022
Lunderberg
added a commit
to Lunderberg/tvm
that referenced
this pull request
Oct 31, 2022
This parameter was introduced in apache#12833, and was passed for all subcommands defined using `generate_command`. However, this broke the `ci.py lint` subcommand. This PR makes the `additional_flags` parameter to the `docker` function be optional, to avoid this breakage.
xinetzone
pushed a commit
to daobook/tvm
that referenced
this pull request
Nov 10, 2022
…che#12833) * [DOCKER][Adreno] Docker infra for Adreno target with CLML support New docker for Adreno that inherits gpu docker and amends android-sdk. For CLML, we need to specify ADRENO_OPENCL via environment variable. CLML SDK can be downlaoded from Qualcomm Developer Network by following https://developer.qualcomm.com/blog/accelerate-your-models-our-opencl-ml-sdk Adreno device is shared with host, hence networking is enabled for Adreno Docker. * * CLML test cases incmpliance with pytest infra * Update tests/scripts/task_build_adreno_bins.sh Co-authored-by: driazati <[email protected]> * * Addional container specific settings via CI invocation * * review comments * * trigger build Co-authored-by: driazati <[email protected]>
xinetzone
pushed a commit
to daobook/tvm
that referenced
this pull request
Nov 10, 2022
…pache#13238) This parameter was introduced in apache#12833, and was passed for all subcommands defined using `generate_command`. However, this broke the `ci.py lint` subcommand. This PR makes the `additional_flags` parameter to the `docker` function be optional, to avoid this breakage.
xinetzone
pushed a commit
to daobook/tvm
that referenced
this pull request
Nov 25, 2022
…che#12833) * [DOCKER][Adreno] Docker infra for Adreno target with CLML support New docker for Adreno that inherits gpu docker and amends android-sdk. For CLML, we need to specify ADRENO_OPENCL via environment variable. CLML SDK can be downlaoded from Qualcomm Developer Network by following https://developer.qualcomm.com/blog/accelerate-your-models-our-opencl-ml-sdk Adreno device is shared with host, hence networking is enabled for Adreno Docker. * * CLML test cases incmpliance with pytest infra * Update tests/scripts/task_build_adreno_bins.sh Co-authored-by: driazati <[email protected]> * * Addional container specific settings via CI invocation * * review comments * * trigger build Co-authored-by: driazati <[email protected]>
xinetzone
pushed a commit
to daobook/tvm
that referenced
this pull request
Nov 25, 2022
…pache#13238) This parameter was introduced in apache#12833, and was passed for all subcommands defined using `generate_command`. However, this broke the `ci.py lint` subcommand. This PR makes the `additional_flags` parameter to the `docker` function be optional, to avoid this breakage.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
New docker for Adreno that inherits gpu docker and amends android-sdk.
For CLML, we need to specify ADRENO_OPENCL via environment variable. CLML SDK can be downlaoded from Qualcomm Developer Network by following
https://developer.qualcomm.com/blog/accelerate-your-models-our-opencl-ml-sdk
Adreno device is shared with host, hence networking is enabled for Adreno Docker.