Skip to content

Remove some code to reduce the size of the mobile inference library.#5252

Merged
hedaoyuan merged 9 commits intoPaddlePaddle:developfrom
hedaoyuan:inference
Nov 14, 2017
Merged

Remove some code to reduce the size of the mobile inference library.#5252
hedaoyuan merged 9 commits intoPaddlePaddle:developfrom
hedaoyuan:inference

Conversation

@hedaoyuan
Copy link
Contributor

@hedaoyuan hedaoyuan commented Oct 31, 2017

The code that removed from the mobile inference library:

  1. SparseRowCpuMatrix
  2. SparseMatrix
  3. Remove some useless layers
Copy link
Contributor

@Xreki Xreki left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have tested this PR in iOS app, the static libraries' size are reduced from

-rw-r--r--  1 liuyiqun01  INTERNAL\Domain Users   8663872 11 14 11:31 libpaddle_capi_engine.a
-rw-r--r--  1 liuyiqun01  INTERNAL\Domain Users  14365784 11 14 11:31 libpaddle_capi_layers.a
-rw-r--r--  1 liuyiqun01  INTERNAL\Domain Users  23029560 11 14 11:31 libpaddle_capi_whole.a

to

-rw-r--r--  1 liuyiqun01  INTERNAL\Domain Users   6863352 11 14 17:30 libpaddle_capi_engine.a
-rw-r--r--  1 liuyiqun01  INTERNAL\Domain Users  10040944 11 14 17:30 libpaddle_capi_layers.a
-rw-r--r--  1 liuyiqun01  INTERNAL\Domain Users  16904200 11 14 17:30 libpaddle_capi_whole.a
@hedaoyuan hedaoyuan merged commit 562599e into PaddlePaddle:develop Nov 14, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

2 participants