User Profile

Posts

Posts

Post not yet marked as solved
4 Replies
0 Views
Replied In coremltools 0.6.2/3 deconvolution slowdown
1) compile model on device2) go to model dir and change the flag "hint_fallback_from_metal" in the file model.espresso.net
Post not yet marked as solved
13 Replies
0 Views
Replied In [Announcement] coremltools-0.6.3 release notes
Thank you for reply1) Yes 2x slowdown on iphone SE2) Previous mlmodel was converted by 0.4.0 because 0.5.1 return error with deconvolution layers. And yes, when I reverted, model performance became as usual. I use 0.4.0 model with Xcode beta (9M136h) and 0.6.2 (try 0.6.3 - same problem) with Xcode beta 6 (9M214v).3) I've performed binary diff of 0.4.0 and 0.6.2 converted models and it was 14 bytes difference
Post not yet marked as solved
13 Replies
0 Views
Replied In [Announcement] coremltools-0.6.3 release notes
After I converted my caffe model with this version, prediction on Iphone SE became 2 time slower. With this version of converter I've started to use Xcode 9 beta 6. May be problem not only with converter 0.6.2.
Post marked as solved
3 Replies
0 Views
Replied In Get from pytorch code to mlmodel
try this https://github.com/prisma-ai/torch2coremlIt seems to will works with pytorch
Post not yet marked as solved
2 Replies
0 Views
Replied In Convert caffe model with PReLU layer
Same error, this made me change PReLU on ReLU.
Post not yet marked as solved
2 Replies
0 Views
Replied In Error when convert caffe eltwise layer
It's ok, each coeff for each bottom. Coremltools parse and load caffe model, this error appears in the end of conversion.