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: content/learning-paths/laptops-and-desktops/win_on_arm_build_onnxruntime/1-dev-env-setup.md
+24-14Lines changed: 24 additions & 14 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -10,29 +10,39 @@ layout: learningpathall
10
10
11
11
In this Learning Path, you'll learn build and deploy a large language model (LLM) on a Windows on Arm (WoA) laptop using ONNX Runtime for inference.
12
12
13
-
You'll first learn how to build the ONNX Runtime and ONNX Runtime Generate() API library and then how to download the Phi-3 model and run the inference. You'll run the short context (4k) mini (3.3B) variant of Phi 3 model. The short context version accepts a shorter (4K) prompts and produces shorter output text compared to the long (128K) context version. The short version consumes less memory.
13
+
You'll learn how to:
14
14
15
-
Your first task is to prepare a development environment with the required software:
15
+
* Build ONNX Runtime and the Generate() API library.
16
+
* Download the Phi-3 model and run inference.
17
+
* Run the short-context (4K) Mini (3.3B) variant of Phi 3 model.
16
18
17
-
- Visual Studio 2022 IDE (latest version recommended)
18
-
- Python 3.10 or higher
19
-
- CMake 3.28 or higher
19
+
{{% notice Note %}}
20
+
The short-context version accepts shorter (4K) prompts and generates shorter outputs than the long-context (128K) version. It also consumes less memory.
21
+
{{% /notice %}}
22
+
23
+
## Set up your Development Environment
24
+
25
+
Your first task is to prepare a development environment with the required software. Start by installing the required tools:
26
+
27
+
- Visual Studio 2022 IDE (latest version recommended).
28
+
- Python 3.10 or higher.
29
+
- CMake 3.28 or higher.
20
30
21
-
The following instructions were tested on a WoA 64-bit Windows machine with at least 16GB of RAM.
31
+
These instructions were tested on a 64-bit WoA machine with at least 16GB of RAM.
22
32
23
-
## Install Visual Studio 2022 IDE
33
+
## Install and Configure Visual Studio 2022
24
34
25
-
Follow these steps to install and configure Visual Studio 2022 IDE:
35
+
Follow these steps:
26
36
27
37
1. Download the latest [Visual Studio IDE](https://visualstudio.microsoft.com/downloads/).
28
38
29
-
2. Select the **Community** edition. An installer called *VisualStudioSetup.exe* will be downloaded.
39
+
2. Select the **Community** edition. This downloads an installer called *VisualStudioSetup.exe*.
30
40
31
-
3. Run the downloaded installer (*VisualStudioSetup.exe*) from your **Downloads** folder.
41
+
3. Run the installer (*VisualStudioSetup.exe*) from your **Downloads** folder.
32
42
33
-
4. Follow the installation prompts and accept the **License Terms** and **Privacy Statement**.
43
+
4. Follow the prompts and accept the **License Terms** and **Privacy Statement**.
34
44
35
-
5. When prompted to select your workloads, select **Desktop Development with C++**. This includes**Microsoft Visual Studio Compiler** (**MSVC**).
45
+
5. When prompted to select workloads, select **Desktop Development with C++**. This installs the**Microsoft Visual Studio Compiler** (**MSVC**).
36
46
37
47
## Install Python
38
48
@@ -44,12 +54,12 @@ You'll need Python version 3.10 or higher. This Learning Path was tested with ve
44
54
45
55
## Install CMake
46
56
47
-
CMake is an open-source tool that automates the build process and helps generate platform-specific build configurations.
57
+
CMake is an open-source tool that automates the build process and generates platform-specific build configurations.
48
58
49
59
Download and install [CMake for Windows on Arm](/install-guides/cmake).
50
60
51
61
{{% notice Note %}}
52
62
The instructions were tested with version 3.30.5.
53
63
{{% /notice %}}
54
64
55
-
You’re now ready to move on to building the ONNX Runtime and running inference with Phi-3.
65
+
You’re now ready to build ONNX Runtime and run inference using the Phi-3 model.
0 commit comments