You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+7-7Lines changed: 7 additions & 7 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -54,7 +54,7 @@ The framework is Alpha testing now. Documents and examples are coming soon.
54
54
### TensorFlow C API Library Installation
55
55
56
56
Perfect-TensorFlow is based on TensorFlow C API, i.e., `libtensorflow.so` on runtime.
57
-
This project contains two express CPU v1.1.0 installation scripts for this module on both macOS / Ubuntu Linux, and will install the dynamic library into path `/usr/local/lib/libtensorflow.so`. You can download & run [`install.mac.sh`](https://github.com/PerfectlySoft/Perfect-TensorFlow/blob/master/install.mac.sh) for macOS X or [`install.linux.sh`](https://github.com/PerfectlySoft/Perfect-TensorFlow/blob/master/install.linux.sh).
57
+
This project contains an express CPU v1.1.0 installation script for this module on both macOS / Ubuntu Linux, and will install the dynamic library into path `/usr/local/lib/libtensorflow.so`. You can download & run [`install.sh`](https://github.com/PerfectlySoft/Perfect-TensorFlow/blob/master/install.sh).
58
58
59
59
For more installation options, such as GPU/CPU and multiple versions on the same machine, please check TensorFlow website: [Installing TensorFlow for C](https://www.tensorflow.org/install/install_c)
let op =try g.const(tensor: tensor, name: "hello")
111
111
@@ -123,7 +123,7 @@ print(s2)
123
123
### Matrix Operations
124
124
125
125
126
-
As you can see, Swift version of TensorFlow keeps the same principals of the original one, i.e., create tensors, save tensors into graph, define the operations and then run the session & check the result.
126
+
As you can see, Swift version of TensorFlow keeps the same principals of the original one, i.e., create tensors, save tensors into graph, define the operations and then run the session & check the result.
127
127
128
128
Here is an other simple example of matrix operations in Perfect TensorFlow:
129
129
@@ -132,8 +132,8 @@ Here is an other simple example of matrix operations in Perfect TensorFlow:
132
132
| 1 2 | |0 1| |0 1|
133
133
| 3 4 | * |0 0| = |0 3|
134
134
*/
135
-
// input the matrix.
136
-
// *NOTE* no matter how many dimensions a matrix may have,
135
+
// input the matrix.
136
+
// *NOTE* no matter how many dimensions a matrix may have,
137
137
// the matrix should always input as an flattened array
138
138
let srcA:[Float] = [[1, 2], [3, 4]].flatMap { $0 }
139
139
let srcB:[Float] = [[0, 0], [1, 0]].flatMap { $0 }
@@ -170,12 +170,12 @@ let metaBuf = try TF.Buffer()
170
170
// load the session
171
171
let session =try g.load(
172
172
exportDir: "/path/to/saved/model",
173
-
tags: ["tag1", "tag2", ...],
173
+
tags: ["tag1", "tag2", ...],
174
174
metaGraphDef: metaBuf)
175
175
```
176
176
177
177
A detailed example of loading model can be found in the [Perfect TensorFlow Testing Examples](https://github.com/PerfectlySoft/Perfect-TensorFlow/blob/master/Tests/PerfectTensorFlowTests/PerfectTensorFlowTests.swift#L349-L390).
178
-
178
+
179
179
## Issues
180
180
181
181
We are transitioning to using JIRA for all bugs and support related issues, therefore the GitHub issues has been disabled.
本项目包含了两个用于快速安装该链接库 CPU v1.1.0版本的脚本,一个用于macOS而另外一个用于Ubuntu Linux。默认安装路径为`/usr/local/lib/libtensorflow.so`。您可以根据平台要求下载并运行 [`install.mac.sh`](https://github.com/PerfectlySoft/Perfect-TensorFlow/blob/master/install.mac.sh)或[`install.linux.sh`](https://github.com/PerfectlySoft/Perfect-TensorFlow/blob/master/install.linux.sh).
58
+
本项目包含了一个用于快速安装该链接库 CPU v1.1.0版本的脚本,默认安装路径为`/usr/local/lib/libtensorflow.so`。您可以根据平台要求下载并运行 [`install.sh`](https://github.com/PerfectlySoft/Perfect-TensorFlow/blob/master/install.sh)。
59
59
60
60
更多的安装选项,如需要在同一台计算机上同时安装CPU/GPU或者多个不同版本,请参考官网网站: [Installing TensorFlow for C](https://www.tensorflow.org/install/install_c)
0 commit comments