Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
nl
on Jan 18, 2023
|
parent
|
context
|
favorite
| on:
Which GPU(s) to get for deep learning
The lack of ECC memory is almost certainly not a factor. If you can train at FP8 your model will recover from a single flipped bit somewhere.
Loranubi
on Jan 19, 2023
[–]
I mean you could even view bit flips as a regularization technique like dropout...
dahart
on Jan 19, 2023
|
parent
[–]
Yeah I hear it’s common practice now to avoid synchronizing GPU training kernels in order to speed things up, and it has positive regularization benefits and little downside.
Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
