projects / platform / core / ml / nnfw.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 2fde53a) | patch
Early version of freezer ( Tensorflow model (code level) to frozen model ) (#419)
author윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 <hyunsik.yoon@samsung.com>
Tue, 10 Apr 2018 06:47:50 +0000 (15:47 +0900)
committer서상민/동작제어Lab(SR)/Senior Engineer/삼성전자 <sangmin7.seo@samsung.com>
Tue, 10 Apr 2018 06:47:50 +0000 (15:47 +0900)
commitc95233b1a4f31b848d9e2b182b9fa5779f1a44ba
treee0d9b0f4bf3ecee750417dba94bf7f62695e93b3tree | snapshot
parent2fde53a1a5d2b1f0d7123e2c55fb730234c473b8commit | diff
Early version of freezer ( Tensorflow model (code level) to frozen model ) (#419)

This commit introduces a tool to convert Tensorflow model (in code level) to frozen model file.
This could be used in future to generate many frozen model files for test cases.
The following samples to generate frozen model are included.

- relu(wx+b)
- softmax(wx+b)
tools/tensorflow_model_freezer/freeze_programmed_tensor_graph.py [new file with mode: 0644] blob
tools/tensorflow_model_freezer/model_freezer_util.py [new file with mode: 0644] blob
tools/tensorflow_model_freezer/readme.md [new file with mode: 0644] blob
Domain: Machine Learning / ML Framework;