Giters

microsoft / MMdnn

MMdnn is a set of tools to help users inter-operate among different deep learning frameworks. E.g. model conversion and visualization. Convert models between Caffe, Keras, MXNet, Tensorflow, CNTK, PyTorch Onnx and CoreML.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

How does MMDNN implement the "retrain" part?

calvin886 opened this issue · comments

commented

Platform (like ubuntu 16.04/win10):

Python version:

Source framework with version (like Tensorflow 1.4.1 with GPU):

Destination framework with version (like CNTK 2.3 with GPU):

Pre-trained model path (webpath or webdisk path):

Running scripts:
Since i saw mmdnn have the ability to "retain" a model, i do not really got how it works.
Does it mean we still can train our model on the mmdnn's IR or what ?

commented

Sorry, didn't quite get your question. i am afraid that one can not train on the IR level, but once the model is transferred to another framework, then you can just use the model anyway you like.