[CustomLayer] Update readme.md
authorJihoon Lee <jhoon.it.lee@samsung.com>
Mon, 30 Nov 2020 07:34:09 +0000 (16:34 +0900)
committerJijoong Moon <jijoong.moon@samsung.com>
Tue, 1 Dec 2020 01:45:27 +0000 (10:45 +0900)
Add readme.md about how to run and expected output

**Self evaluation:**
1. Build test: [X]Passed [ ]Failed [ ]Skipped
2. Run test: [ ]Passed [ ]Failed [X]Skipped

Signed-off-by: Jihoon Lee <jhoon.it.lee@samsung.com>
Applications/Custom/LayerClient/README.md

index 555a1c2..929ca69 100644 (file)
@@ -2,3 +2,54 @@
 
 This application demonstrates how to create a custom layer and register inside client-side code
 
+# How to
+
+1. build with ninja
+
+2. run with ini path or argument `model`
+
+```bash
+$ Applications/Custom/LayerClient/jni/layer_client model
+pow layer forward is called
+input: <N9nntrainer6TensorE at 0x55d4521c75e0>
+Shape: 10:1:1:100
+[2 2 2 ... 2 2 2]
+output: <N9nntrainer6TensorE at 0x55d4521cb7f0>
+Shape: 10:1:1:100
+[8 8 8 ... 8 8 8]
+Press enter key to continue...
+pow layer backward is called
+input: <N9nntrainer6TensorE at 0x55d4521cb7f0>
+Shape: 10:1:1:100
+[0.00556926 0.000204645 -0.00607976 ... 0.022363 0.020753 0.00855796]
+output: <N9nntrainer6TensorE at 0x55d4521c75e0>
+Shape: 10:1:1:100
+[0.0167078 0.000613935 -0.0182393 ... 0.067089 0.0622591 0.0256739]
+Press enter key to continue...
+pow layer forward is called ]  ( Training Loss: 0 )
+input: <N9nntrainer6TensorE at 0x55d4521c75e0>
+Shape: 10:1:1:100
+[2 2 2 ... 2 2 2]
+output: <N9nntrainer6TensorE at 0x55d4521cb7f0>
+Shape: 10:1:1:100
+[8 8 8 ... 8 8 8]
+Press enter key to continue...
+
+pow layer backward is called
+input: <N9nntrainer6TensorE at 0x55d4521cb7f0>
+Shape: 10:1:1:100
+[0.0045707 -0.00549962 0.014237 ... -0.0178485 -0.0163203 -0.00993024]
+output: <N9nntrainer6TensorE at 0x55d4521c75e0>
+Shape: 10:1:1:100
+[0.0137121 -0.0164989 0.0427111 ... -0.0535455 -0.0489608 -0.0297907]
+Press enter key to continue...
+
+pow layer forward is called ]  ( Training Loss: 0 )
+input: <N9nntrainer6TensorE at 0x55d4521c75e0>
+Shape: 10:1:1:100
+[2 2 2 ... 2 2 2]
+output: <N9nntrainer6TensorE at 0x55d4521cb7f0>
+Shape: 10:1:1:100
+[8 8 8 ... 8 8 8]
+Press enter key to continue...
+```