self-hosted/immich
1
0
Fork 0
mirror of https://github.com/immich-app/immich.git synced 2024-12-26 10:50:29 +02:00
Code Issues Releases Activity

docs(ml): clarify cuda and openvino requirements (#9072)

Browse Source 
update docs
This commit is contained in:
Mert 2024-04-25 00:15:26 -04:00 committed by GitHub
parent 466451abc9
commit a90138e42e
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 10 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
docs/docs/features/ml-hardware-acceleration.md
View File

@ -10,15 +10,14 @@ You do not need to redo any machine learning jobs after enabling hardware accele
## Supported Backends ## Supported Backends
- ARM NN (Mali) - ARM NN (Mali)
- CUDA (NVIDIA) Note: It is supported with [compute capability](https://developer.nvidia.com/cuda-gpus) 5.2 or higher - CUDA (NVIDIA GPUs with [compute capability](https://developer.nvidia.com/cuda-gpus) 5.2 or higher)
- OpenVINO (Intel) - OpenVINO (Intel discrete GPUs such as Iris Xe and Arc)
## Limitations ## Limitations
- The instructions and configurations here are specific to Docker Compose. Other container engines may require different configuration. - The instructions and configurations here are specific to Docker Compose. Other container engines may require different configuration.
- Only Linux and Windows (through WSL2) servers are supported. - Only Linux and Windows (through WSL2) servers are supported.
- ARM NN is only supported on devices with Mali GPUs. Other Arm devices are not supported. - ARM NN is only supported on devices with Mali GPUs. Other Arm devices are not supported.
- There is currently an upstream issue with OpenVINO, so whether it will work is device-dependent.
- Some models may not be compatible with certain backends. CUDA is the most reliable. - Some models may not be compatible with certain backends. CUDA is the most reliable.
## Prerequisites ## Prerequisites
@ -36,9 +35,16 @@ You do not need to redo any machine learning jobs after enabling hardware accele
#### CUDA #### CUDA
- You must have the official NVIDIA driver installed on the server. - The GPU must have compute capability 5.2 or greater.
- The server must have the official NVIDIA driver installed.
- The installed driver must be >= 535 (it must support CUDA 12.2).
- On Linux (except for WSL2), you also need to have [NVIDIA Container Runtime][nvcr] installed. - On Linux (except for WSL2), you also need to have [NVIDIA Container Runtime][nvcr] installed.
#### OpenVINO
- The server must have a discrete GPU, i.e. Iris Xe or Arc. Expect issues when attempting to use integrated graphics.
- Ensure the server's kernel version is new enough to use the device for hardware accceleration.
## Setup ## Setup
1. If you do not already have it, download the latest [`hwaccel.ml.yml`][hw-file] file and ensure it's in the same folder as the `docker-compose.yml`. 1. If you do not already have it, download the latest [`hwaccel.ml.yml`][hw-file] file and ensure it's in the same folder as the `docker-compose.yml`.