Unraid Community Apps templates by Aurial

Two templates live in this repository: Strata AMD (the AMD/ROCm builds of Strata, a local LLM) and Phonon TTS (the official speech server for Gradium's Phonon text-to-speech models).

Strata AMD

An Unraid Community Applications template for the AMD (ROCm) builds of Strata, which runs the Qwen3.8-Flash-Next mixture-of-experts model on an AMD GPU plus system RAM. It serves an OpenAI- and Anthropic-compatible API, a chat page and a live monitor.

The template runs Strata's official image, ghcr.io/niko1221/strata:latest-rocm-gfx120x. Report template or Unraid problems in this repository's issues, and problems with Strata itself to the Strata project.

Why a separate listing

The Community Apps catalog already lists Strata, an NVIDIA-only template by williamkunz (CUDA images, Nvidia Driver plugin, --runtime=nvidia). That template cannot run on AMD cards. This listing covers the AMD side of the same upstream project: different image flavors (rocm-gfx120x / rocm-gfx110x / rocm-gfx1151), different host requirements (the amdgpu kernel driver Unraid already ships — no driver plugin), and device passthrough via /dev/kfd and /dev/dri instead of the NVIDIA runtime. Both templates use Strata's official image and expose the same settings.

Install

Search for Strata AMD in the Apps tab. Before installing you need:

• an AMD card from one of the families Strata publishes images for. At install, pick your card under Branch — the template offers all three ROCm flavors as a dropdown:

  Branch shown at install	Card
  latest-rocm-gfx120x (default)	RX 9000 series, Radeon AI PRO R9700 (RDNA 4)
  latest-rocm-gfx110x	RX 7000 series (RDNA 3)
  latest-rocm-gfx1151	Ryzen AI Max "Strix Halo" (Radeon 8060S)

  The RX 6000 series has no image yet. The host needs only the amdgpu kernel driver (Linux 6.8+), which Unraid ships as standard — no driver plugin to install.

• 32 GB of RAM or more; more RAM allows a larger, better model (Strata's model guide)

• room for the model on an SSD pool, mapped as a direct pool path (/mnt/cache/...), not /mnt/user

Do not set a memory limit on the container: Strata plans from the server's whole RAM. Do not pin CPUs: the engine sets its own thread affinity.

The first start downloads the model and prepares it. Later starts go straight to serving. The template's Overview explains each setting, and Strata's Docker guide covers them in full.

An API key is always on. If you leave it empty, the first start makes one; show it with:

docker exec "Strata AMD" strata launch --show-key --preset docker

(The container name contains a space, so quote it.)

Multi-GPU

The GPUs field takes the card numbers as the monitor lists them (0,1 splits the model across both by layer parallelism), or all / empty for the engine's choice. Every named card must match the Branch family. Under Advanced there is a GPU roles field for finer control (1=main,0=helper): main runs the final stage the answer comes from, helper holds a share of the model's layers, off leaves a card alone; cards you do not name are off, and Strata's automatic choice (faster card as main) is right for almost everyone. A split adds capacity more than speed — see Strata's multi-GPU guide.

Updates

The template follows latest-rocm-gfx120x, which Strata's maintainer moves to each new release when its image is published; the same holds for the other two Branch tags. Unraid shows the update on the Docker tab for whichever branch you installed. To stay on one release, change the Repository tag to a version such as 0.2.7-rocm-gfx120x.

Phonon TTS

An Unraid Community Applications template for ptts-openai-server, the official speech server for Phonon, Gradium's open-weight on-device text-to-speech models. It serves an OpenAI-compatible API (POST /v1/audio/speech) and the Wyoming protocol, so it plugs into Open WebUI, SillyTavern, Home Assistant and anything else with an "OpenAI TTS" setting — fully local, CPU-only, and fast enough to speak faster than real time even on a Raspberry Pi.

Search for Phonon TTS in the Apps tab; the defaults work out of the box: API port 8880, language en, and a model/voices cache (under about 1 GB) downloaded on first start and kept across restarts. The first start logs little until the model loads — watch the log. For Home Assistant, set Wyoming address to 0.0.0.0:10200, map the Wyoming port, then add the Wyoming Protocol integration pointing at this server. The template's Overview explains every setting, including other models (hf://owner/name), gated 40M checkpoints, voice aliases and OpenAI voice-name compatibility.

The server has no authentication on any port. Run it on a trusted network; never expose it to the internet.

The template follows latest, which Gradium moves with each release; pick a pinned tag (such as 0.5.0) under Branch at install to stay on one release.

Credits and licence

Strata and its images are by Niko1221 and the Strata contributors, under the MIT licence. The Strata icon is Strata's own. The Strata template follows the structure of the NVIDIA strata-docker template by William Kunz (MIT).

Phonon and the xn-ptts server are by Gradium and contributors, Apache-2.0; the models themselves carry their own licences on Hugging Face (the 31M models are Apache-2.0; the 40M models are gated — check their model cards). The Phonon icon in this repository is an original work by the template maintainer, MIT.

This repository is MIT licensed; see LICENSE.