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
+31-11Lines changed: 31 additions & 11 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -15,24 +15,34 @@ ImgSearch is a lightweight image search engine that supports image-to-image and
15
15
16
16
## Installation
17
17
18
-
### Standard Installation
18
+
### Full Installation (Client + Server)
19
19
20
-
ImgSearch relies on TinyCLIP models (built on OpenCLIP), which require PyTorch. The standard installation automatically pulls in the CUDA version of PyTorch. This is recommended for users with NVIDIA GPUs to speed up image inference.
20
+
To use the full functionality of ImgSearch (including client and server), you need to specify the `all` dependency group during installation. This will install all the dependency packages required to run the server, such as `PyTorch`, on the system.
21
+
22
+
```shell
23
+
pip install 'imgsearch[all]'
24
+
```
25
+
26
+
### Standard Installation (Client Only)
27
+
28
+
The standard installation includes only the client-side dependencies:
21
29
22
30
```shell
23
31
pip install imgsearch
24
32
```
25
33
26
-
### CPU Environment
34
+
**⚠️ Note**: Since the standard installation lacks PyTorch and other dependencies, you cannot run the `isearch service start` command.
35
+
36
+
### Non-CUDA environment
27
37
28
-
For non-CUDA environments, install the CPU version of PyTorch first, then install ImgSearch:
38
+
For users in non-CUDA environments who want to run the server, it is recommended to install the CPU version of PyTorch first, and then install ImgSearch:
0 commit comments