homelab/active/software_ai_stack/ai_stack.md

# Self Hosted AI Stack

- [Self Hosted AI Stack](#self-hosted-ai-stack)
  - [Notes](#notes)
    - [Podman Volume Locations](#podman-volume-locations)
  - [Setup](#setup)
    - [Create the AI user](#create-the-ai-user)
    - [Helper aliases](#helper-aliases)
    - [Create the models dir](#create-the-models-dir)
    - [Install the Hugging Face CLI](#install-the-hugging-face-cli)
    - [Samba Model Storage](#samba-model-storage)
    - [Download models](#download-models)
      - [Text models](#text-models)
        - [GPT-OSS](#gpt-oss)
        - [Mistral](#mistral)
        - [Qwen](#qwen)
        - [GLM](#glm)
        - [Gemma](#gemma)
        - [Dolphin](#dolphin)
      - [Image models](#image-models)
        - [Z-Image](#z-image)
        - [Flux](#flux)
      - [Embedding Models](#embedding-models)
        - [Nomic](#nomic)
  - [llama.cpp](#llamacpp)
  - [stable-diffusion.cpp](#stable-diffusioncpp)
  - [open-webui](#open-webui)
  - [Install Services with Quadlets](#install-services-with-quadlets)
    - [Internal and External Pods](#internal-and-external-pods)
    - [Llama CPP Server](#llama-cpp-server)
    - [Stable Diffusion CPP](#stable-diffusion-cpp)
    - [Open Webui](#open-webui-1)
    - [Install the update script](#install-the-update-script)
    - [Install Guest Open Webui with Start/Stop Services](#install-guest-open-webui-with-startstop-services)
  - [Benchmark Results](#benchmark-results)

## Notes

### Podman Volume Locations

`~/.local/share/containers/storage/volumes/`

## Setup

### Create the AI user

```bash
# Create your local ai user. This will be the user you launch podman processes from.
useradd -m ai
loginctl enable-linger ai
su -l ai
mkdir -p /home/ai/.config/containers/systemd/
mkdir -p /home/ai/.ssh
```

Models are big. You'll want some tools to help find large files quickly when space runs out.

### Helper aliases

Add these to your .bashrc:

```bash
# Calculate all folder sizes in current dir 
alias {dudir,dud}='du -h --max-depth 1 | sort -h'

# Calculate all file sizes in current dir
alias {dufile,duf}='ls -lhSr'

# Restart llama-server / follow logs
alias llama-reload="systemctl --user daemon-reload && systemctl --user restart llama-server.service"
alias llama-logs="journalctl --user -fu llama-server"

# Restart stable diffusion gen and edit server / follow logs
alias sd-gen-reload='systemctl --user daemon-reload && systemctl --user restart stable-diffusion-gen-server'
alias sd-gen-logs='journalctl --user -xeu stable-diffusion-gen-server'
alias sd-edit-reload='systemctl --user daemon-reload && systemctl --user restart stable-diffusion-edit-server'
alias sd-edit-logs='journalctl --user -xeu stable-diffusion-edit-server'
```

### Create the models dir

```bash
mkdir -p /home/ai/models/{text,image,video,embedding,tts,stt}
```

### Install the Hugging Face CLI

<https://huggingface.co/docs/huggingface_hub/en/guides/cli#getting-started>

```bash
# Install
curl -LsSf https://hf.co/cli/install.sh | bash

# Login
hf auth login
```

### Samba Model Storage

I recommend adding network storage for keeping models offloaded. This mounts a samba share at `/srv/models`.

```bash
dnf install -y cifs-utils

# Add this to /etc/fstab
//driveripper.reeselink.com/smb_models /srv/models cifs _netdev,nofail,uid=1001,gid=1001,credentials=/etc/samba/credentials 0 0

# Then mount
systemctl daemon-reload
mount -a --mkdir
```

Here are some sync commands that I use to keep the samba share in sync with the home directory:

```bash
# Sync models from home dir to the samba share
rsync -av --progress /home/ai/models/ /srv/models/
```

### Download models

In general I try to run 8 bit quantized minimum.

#### Text models

<https://huggingface.co/ggml-org/collections>

##### GPT-OSS

<https://unsloth.ai/docs/models/gpt-oss-how-to-run-and-fine-tune#recommended-settings>

```bash
# gpt-oss-120b
mkdir gpt-oss-120b && cd gpt-oss-120b
hf download --local-dir . ggml-org/gpt-oss-120b-GGUF

# gpt-oss-20b
mkdir gpt-oss-20b && cd gpt-oss-20b
hf download --local-dir . ggml-org/gpt-oss-20b-GGUF
```

##### Mistral

```bash
# devstral-small-2-24b
mkdir devstral-small-2-24b && cd devstral-small-2-24b
hf download --local-dir . ggml-org/Devstral-Small-2-24B-Instruct-2512-GGUF Devstral-Small-2-24B-Instruct-2512-Q8_0.gguf

# ministral-3-14b
mkdir ministral-3-14b && cd ministral-3-14b
hf download --local-dir . ggml-org/Ministral-3-14B-Reasoning-2512-GGUF

# ministral-3-3b-instruct
mkdir ministral-3-3b-instruct && cd ministral-3-3b-instruct
hf download --local-dir . ggml-org/Ministral-3-3B-Instruct-2512-GGUF
```

##### Qwen

```bash
# qwen3-30b-a3b-thinking
mkdir qwen3-30b-a3b-thinking && cd qwen3-30b-a3b-thinking
hf download --local-dir . ggml-org/Qwen3-30B-A3B-Thinking-2507-Q8_0-GGUF

# qwen3-30b-a3b-instruct
mkdir qwen3-30b-a3b-instruct && cd qwen3-30b-a3b-instruct
hf download --local-dir . ggml-org/Qwen3-30B-A3B-Instruct-2507-Q8_0-GGUF

# qwen3-vl-30b-a3b-thinking
mkdir qwen3-vl-30b-a3b-thinking && cd qwen3-vl-30b-a3b-thinking
hf download --local-dir . Qwen/Qwen3-VL-30B-A3B-Thinking-GGUF Qwen3VL-30B-A3B-Thinking-Q8_0.gguf
hf download --local-dir . Qwen/Qwen3-VL-30B-A3B-Thinking-GGUF mmproj-Qwen3VL-30B-A3B-Thinking-F16.gguf

# qwen3-vl-30b-a3b-instruct
mkdir qwen3-vl-30b-a3b-instruct && cd qwen3-vl-30b-a3b-instruct
hf download --local-dir . Qwen/Qwen3-VL-30B-A3B-Instruct-GGUF Qwen3VL-30B-A3B-Instruct-Q8_0.gguf
hf download --local-dir . Qwen/Qwen3-VL-30B-A3B-Instruct-GGUF mmproj-Qwen3VL-30B-A3B-Instruct-F16.gguf

# qwen3-coder-30b-a3b-instruct
mkdir qwen3-coder-30b-a3b-instruct && cd qwen3-coder-30b-a3b-instruct
hf download --local-dir . ggml-org/Qwen3-Coder-30B-A3B-Instruct-Q8_0-GGUF

# qwen3-coder-next
mkdir qwen3-coder-next && cd qwen3-coder-next
hf download --local-dir . unsloth/Qwen3-Coder-Next-GGUF --include "Q8_0/*.gguf"
```

##### GLM

```bash
# glm-4.7-flash-30b
mkdir glm-4.7-flash-30b && cd glm-4.7-flash-30b
hf download --local-dir . unsloth/GLM-4.7-Flash-GGUF GLM-4.7-Flash-Q8_0.gguf
```

##### Gemma

```bash
# Note "it" vs "pt" suffixes. "it" is instruction following, "pt" is the base model (not as good for out-of-the-box use)
# gemma-3-27b-it
mkdir gemma-3-27b-it && cd gemma-3-27b-it
hf download --local-dir . unsloth/gemma-3-27b-it-GGUF gemma-3-27b-it-Q8_0.gguf
hf download --local-dir . unsloth/gemma-3-27b-it-GGUF mmproj-F16.gguf
```

##### Dolphin

```bash
# dolphin-mistral-24b-venice
mkdir dolphin-mistral-24b-venice && cd dolphin-mistral-24b-venice
cd dolphin-mistral-24b-venice
hf download --local-dir . bartowski/cognitivecomputations_Dolphin-Mistral-24B-Venice-Edition-GGUF cognitivecomputations_Dolphin-Mistral-24B-Venice-Edition-Q8_0.gguf
```

#### Image models

##### Z-Image

```bash
# z-turbo
# Fastest image generation in 8 steps. Great a text and prompt following.
# Lacks variety.
mkdir /home/ai/models/image/z-turbo && cd /home/ai/models/image/z-turbo
hf download --local-dir . leejet/Z-Image-Turbo-GGUF z_image_turbo-Q8_0.gguf
hf download --local-dir . black-forest-labs/FLUX.1-schnell ae.safetensors
hf download --local-dir . unsloth/Qwen3-4B-Instruct-2507-GGUF Qwen3-4B-Instruct-2507-Q8_0.gguf
```

##### Flux

```bash
# flux2-klein
# Capable of editing images in 4 steps (though 5 is my recommended steps)
mkdir /home/ai/models/image/flux2-klein && cd /home/ai/models/image/flux2-klein
hf download --local-dir . leejet/FLUX.2-klein-9B-GGUF flux-2-klein-9b-Q8_0.gguf
hf download --local-dir . black-forest-labs/FLUX.2-dev ae.safetensors
hf download --local-dir . unsloth/Qwen3-8B-GGUF Qwen3-8B-Q8_0.gguf
```

#### Embedding Models

##### Nomic

```bash
# nomic-embed-text-v2
mkdir /home/ai/models/embedding/nomic-embed-text-v2
hf download --local-dir /home/ai/models/embedding/nomic-embed-text-v2 ggml-org/Nomic-Embed-Text-V2-GGUF
```

## llama.cpp

<https://github.com/ggml-org/llama.cpp/tree/master/tools/server>

```bash
# Build the llama.cpp container image
git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
export BUILD_TAG=$(date +"%Y-%m-%d-%H-%M-%S")

# Vulkan (better performance as of Feb 2026)
podman build -f .devops/vulkan.Dockerfile -t llama-cpp-vulkan:${BUILD_TAG} -t llama-cpp-vulkan:latest .

# ROCM
podman build -f .devops/rocm.Dockerfile -t llama-cpp-rocm:${BUILD_TAG} -t llama-cpp-rocm:latest .

# Run llama demo server (Available on port 8000)
podman run \
--rm \
--name llama-server-demo \
--device=/dev/kfd \
--device=/dev/dri \
-v /home/ai/models/text:/models:z \
-p 8000:8000 \
localhost/llama-cpp-vulkan:latest \
--host 0.0.0.0 \
--port 8000 \
-c 32768 \
--perf \
--n-gpu-layers all \
--jinja \
--models-max 1 \
--models-dir /models
```

## stable-diffusion.cpp

Server: <https://github.com/leejet/stable-diffusion.cpp/tree/master/examples/server>

CLI: <https://github.com/leejet/stable-diffusion.cpp/tree/master/examples/cli>

```bash
git clone https://github.com/leejet/stable-diffusion.cpp.git
cd stable-diffusion.cpp
git submodule update --init --recursive
export BUILD_TAG=$(date +"%Y-%m-%d-%H-%M-%S")

# Vulkan
podman build -f Dockerfile.vulkan -t stable-diffusion-cpp:${BUILD_TAG} -t stable-diffusion-cpp:latest .
```

```bash
# Generate an image with z-turbo
podman run --rm \
-v /home/ai/models:/models:z \
-v /home/ai/output:/output:z \
--device /dev/kfd \
--device /dev/dri \
localhost/stable-diffusion-cpp:latest \
--diffusion-model /models/image/z-turbo/z_image_turbo-Q8_0.gguf \
--vae /models/image/z-turbo/ae.safetensors  \
--llm /models/image/z-turbo/Qwen3-4B-Instruct-2507-Q8_0.gguf \
-v \
--cfg-scale 1.0 \
--vae-conv-direct \
--diffusion-conv-direct \
--fa \
--mmap \
--seed -1 \
--steps 8 \
-H 1024 \
-W 1024 \
-o /output/output.png \
-p "A photorealistic dragon"

# Edit the generated image with flux2-klein
podman run --rm \
-v /home/ai/models:/models:z \
-v /home/ai/output:/output:z \
--device /dev/kfd \
--device /dev/dri \
localhost/stable-diffusion-cpp:latest \
--diffusion-model  /models/image/flux2-klein/flux-2-klein-9b-Q8_0.gguf \
--vae /models/image/flux2-klein/ae.safetensors \
--llm /models/image/flux2-klein/Qwen3-8B-Q8_0.gguf \
-v \
--cfg-scale 1.0 \
--sampling-method euler \
--vae-conv-direct \
--diffusion-conv-direct \
--fa \
--mmap \
--steps 5 \
-H 1024 \
-W 1024 \
-r /output/output.png \
-o /output/edit.png \
-p "Replace the dragon with an old car"
```

## open-webui

```bash
mkdir /home/ai/.env
# Create a file called open-webui-env with `WEBUI_SECRET_KEY="some-random-key"
scp active/device_framework_desktop/secrets/open-webui-env deskwork-ai:.env/

# Will be available on port 8080
podman run \
-d \
-p 8080:8080 \
-v open-webui:/app/backend/data \
--name open-webui \
--restart always \
ghcr.io/open-webui/open-webui:main
```

Use the following connections:

| Service                   | Endpoint                                  |
| ------------------------- | ----------------------------------------- |
| llama.cpp                 | <http://host.containers.internal:8000>    |
| stable-diffusion.cpp      | <http://host.containers.internal:1234/v1> |
| stable-diffusion.cpp edit | <http://host.containers.internal:1235/v1> |

## Install Services with Quadlets

### Internal and External Pods

These will be used to restrict internet access to our llama.cpp and
stable-diffusion.cpp services while allowing the frontend services to
communicate with those containers.

```bash
scp -r active/device_framework_desktop/quadlets_pods/* deskwork-ai:.config/containers/systemd/
ssh deskwork-ai
systemctl --user daemon-reload
systemctl --user start ai-internal-pod.service ai-external-pod.service
```

### Llama CPP Server

Installs the llama.cpp server to run our text models.

```bash
scp -r active/device_framework_desktop/quadlets_llama_server/* deskwork-ai:.config/containers/systemd/
ssh deskwork-ai
systemctl --user daemon-reload
systemctl --user restart ai-internal-pod.service
```

### Stable Diffusion CPP

Installs the stable-diffusion.cpp server to run our image models.

```bash
scp -r active/device_framework_desktop/quadlets_stable_diffusion/* deskwork-ai:.config/containers/systemd/
ssh deskwork-ai
systemctl --user daemon-reload
systemctl --user restart ai-internal-pod.service
```

### Open Webui

Installs the open webui frontend.

```bash
scp -r active/device_framework_desktop/quadlets_openwebui/* deskwork-ai:.config/containers/systemd/
ssh deskwork-ai
systemctl --user daemon-reload
systemctl --user restart ai-external-pod.service
```

Note, all services will be available at `host.containers.internal`. So llama.cpp
will be up at `http://host.containers.internal:8000`.

### Install the update script

```bash
# 1. Builds the latest llama.cpp and stable-diffusion.cpp
# 2. Pulls the latest open-webui
# 3. Restarts all services
scp active/device_framework_desktop/update-script.sh deskwork-ai:
ssh deskwork-ai
chmod +x update-script.sh
./update-script.sh
```

### Install Guest Open Webui with Start/Stop Services

Optionally install a guest openwebui service.

```bash
scp -r active/device_framework_desktop/systemd/. deskwork-ai:.config/systemd/user/
ssh deskwork-ai
systemctl --user daemon-reload
systemctl --user enable open-webui-guest-start.timer
systemctl --user enable open-webui-guest-stop.timer
```

## Benchmark Results

Benchmarks are run with [unsloth gpt-oss-20b Q8_0](https://huggingface.co/unsloth/gpt-oss-20b-GGUF/blob/main/gpt-oss-20b-Q8_0.gguf)

```bash
# Run the llama.cpp pod (AMD)
podman run -it --rm \
--device=/dev/kfd \
--device=/dev/dri \
-v /home/ai/models/text:/models:z \
--entrypoint /bin/bash \
ghcr.io/ggml-org/llama.cpp:full-vulkan

# Benchmark command
./llama-bench -m /models/benchmark/gpt-oss-20b-Q8_0.gguf
```

Framework Desktop

| model            |      size |  params | backend |  ngl |  test |            t/s |
| ---------------- | --------: | ------: | ------- | ---: | ----: | -------------: |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | Vulkan  |   99 | pp512 | 1128.50 ± 7.60 |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | Vulkan  |   99 | tg128 |   77.94 ± 0.08 |

| model            |      size |  params | backend |  ngl |  test |           t/s |
| ---------------- | --------: | ------: | ------- | ---: | ----: | ------------: |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | ROCm    |   99 | pp512 | 526.05 ± 7.04 |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | ROCm    |   99 | tg128 |  70.98 ± 0.01 |

AMD R9700

| model            |      size |  params | backend |  ngl |  test |              t/s |
| ---------------- | --------: | ------: | ------- | ---: | ----: | ---------------: |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | Vulkan  |   99 | pp512 | 3756.79 ± 203.97 |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | Vulkan  |   99 | tg128 |    174.24 ± 0.32 |

NVIDIA GeForce RTX 4080 SUPER

| model            |      size |  params | backend |  ngl |  test |           t/s |
| ---------------- | --------: | ------: | ------- | ---: | ----: | ------------: |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | CUDA    |   99 | tg128 | 193.28 ± 1.03 |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | CUDA    |   99 | tg256 | 193.55 ± 0.34 |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | CUDA    |   99 | tg512 | 187.39 ± 0.10 |

NVIDIA GeForce RTX 3090

| model            |      size |  params | backend |  ngl |  test |             t/s |
| ---------------- | --------: | ------: | ------- | ---: | ----: | --------------: |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | CUDA    |   99 | pp512 | 4297.72 ± 35.60 |
| gpt-oss 20B Q8_0 | 11.27 GiB | 20.91 B | CUDA    |   99 | tg128 |   197.73 ± 0.62 |