-
Notifications
You must be signed in to change notification settings - Fork 61
{2023.06}[2023a,a64fx] add TensorFlow 2.13 #1034
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
{2023.06}[2023a,a64fx] add TensorFlow 2.13 #1034
Conversation
Instance
|
Instance
|
Instance
|
Instance
|
Instance
|
Instance
|
Instance
|
bot: build instance:eessi-bot-deucalion repository:eessi.io-2023.06-software architecture:aarch64/a64fx |
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
New job on instance
|
First job seems a bit slow. Launch one with parallel = 8... |
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
New job on instance
|
Updates by the bot instance
|
Updates by the bot instance
|
New job on instance
|
One more with parallel = 16... |
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
New job on instance
|
…-layer into 2023.06-a64fx-2023a-eb482-apps-tf kept original order to list TensorFlow right after all its dependencies
if cpu_target == CPU_TARGET_A64FX and self.name in ['TensorFlow']: | ||
# limit parallelism to 8, builds with 12 and 16 failed on Deucalion | ||
if parallel > 8: | ||
self.cfg['parallel'] = 8 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@trz42 Why don't we simply use a factor of 4
when building for A64FX, rather than a factor of 2 like we do below?
In theory, we could have smaller build jobs (say with 4 cores) at some point, so really hardcoding to 8
seems wrong to me...
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ah, I overlooked the > 8
condition, sorry
Label |
staging PR merged, so merging this too... |
77651f1
into
EESSI:2023.06-software.eessi.io
PR merged! Moved |
1 similar comment
PR merged! Moved |
PR merged! Moved |
PR merged! Moved |
Adds TensorFlow 2.13. Limits parallelism to 8 (via
eb_hooks.py
) in order to work around out-of-memory issue.