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/4-run-benchmark-on-WoA.md
+11-7Lines changed: 11 additions & 7 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -8,31 +8,35 @@ layout: learningpathall
8
8
9
9
## Run the Phi-3 model on your Windows on Arm machine
10
10
11
-
In this section, you'll download the Phi-3 Mini model and run it on your Windows on Arm machine, which can be a physical or virtual machine. You'll be use a simple model runner program which provides performance metrics
11
+
In this section, you'll download the Phi-3 Mini model and run it on your WoA machine - either physical or virtual. You'll use a simple model runner that also reports performance metrics.
12
12
13
-
The Phi-3 Mini (3.3B) model has a short (4k) context version and a long (128k) context version. The long context version can accept much longer prompts and produces longer output text, but it consumes more memory.
13
+
The Phi-3 Mini (3.3B) model is available in two versions:
14
14
15
-
In this learning path, you'll use the short context version, which is quantized to 4-bits.
15
+
- Short context (4K) - supports shorter prompts and uses less memory.
16
+
- Long context (128K) - supports longer prompts and outputs but consumes more memory.
17
+
18
+
This Learning Path uses the short context version, which is quantized to 4-bits.
16
19
17
20
The Phi-3 Mini model used here is in ONNX format.
18
21
19
22
### Setup
20
23
21
24
[Phi-3 ONNX models](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct-onnx) are hosted on HuggingFace.
22
-
Hugging Face uses Git for both version control and to download the ONNX model files, which can be quite large.
25
+
Hugging Face uses Git for both version control and to download the ONNX model files, which are large.
26
+
27
+
### Install Git LFS
23
28
24
29
You'll first need to install the Git Large File Storage (LFS) extension:
25
30
26
31
```bash
27
32
winget install -e --id GitHub.GitLFS
28
33
git lfs install
29
34
```
30
-
If you don’t have winget, download and run the exe from the [official source](https://docs.github.com/en/repositories/working-with-files/managing-large-files/installing-git-large-file-storage?platform=windows).
35
+
If you don’t have winget, [download the installer manually](https://docs.github.com/en/repositories/working-with-files/managing-large-files/installing-git-large-file-storage?platform=windows).
31
36
32
37
If the extension is already installed for you when you run the above ``git`` command it will say ``Git LFS initialized``.
0 commit comments