Search
Search titles only
By:
Search titles only
By:
Forums
New posts
Search forums
What's new
New posts
Latest activity
Feedback
View Statistics
Members
Current visitors
Buy Sell Trade
WTB
Log in
Register
Search
Search titles only
By:
Search titles only
By:
New posts
Search forums
Menu
Install the app
Install
Reply to thread
Forums
Technology
Programming
Need help with pytorch....cannot resume training
JavaScript is disabled. For a better experience, please enable JavaScript in your browser before proceeding.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an
alternative browser
.
Message
<blockquote data-quote="bhp91" data-source="post: 2465505" data-attributes="member: 77047"><p>That's because GPU's don't have native support for bfloat16. Bfloat16 was developed by Google specifically for their TPU's.</p><p></p><p>You can still get these much smaller model sizes and faster inference by applying <a href="https://www.tensorflow.org/lite/performance/post_training_quantization" target="_blank">post training quantization</a> on a already trained model.</p></blockquote><p></p>
[QUOTE="bhp91, post: 2465505, member: 77047"] That's because GPU's don't have native support for bfloat16. Bfloat16 was developed by Google specifically for their TPU's. You can still get these much smaller model sizes and faster inference by applying [URL='https://www.tensorflow.org/lite/performance/post_training_quantization']post training quantization[/URL] on a already trained model. [/QUOTE]
Insert quotes…
Verification
Post reply
Forums
Technology
Programming
Need help with pytorch....cannot resume training
Top
Bottom