Instructions to use DKTechin/acestep with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use DKTechin/acestep with ACE-Step:

# No code snippets available yet for this library.

# To use this model, check the repository files and the library's documentation.

# Want to help? PRs adding snippets are welcome at:
# https://github.com/huggingface/huggingface.js

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use DKTechin/acestep with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf DKTechin/acestep:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf DKTechin/acestep:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf DKTechin/acestep:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf DKTechin/acestep:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf DKTechin/acestep:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf DKTechin/acestep:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf DKTechin/acestep:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf DKTechin/acestep:Q4_K_M

Use Docker

docker model run hf.co/DKTechin/acestep:Q4_K_M

LM Studio
Jan
Ollama
How to use DKTechin/acestep with Ollama:
```
ollama run hf.co/DKTechin/acestep:Q4_K_M
```

Unsloth Studio

How to use DKTechin/acestep with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for DKTechin/acestep to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for DKTechin/acestep to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for DKTechin/acestep to start chatting

Docker Model Runner
How to use DKTechin/acestep with Docker Model Runner:
```
docker model run hf.co/DKTechin/acestep:Q4_K_M
```

Lemonade

How to use DKTechin/acestep with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull DKTechin/acestep:Q4_K_M

Run and chat with the model

lemonade run user.acestep-Q4_K_M

List all available models

lemonade list

Atomic Chat

acestep / README.md

succeun

Add model card: runtime builds + mirrored ACE-Step 1.5 GGUF weights

b7db683 verified 9 days ago

preview code

Raw

History Blame Contribute Delete

2.48 kB

	---
	license: mit
	base_model:
	- ACE-Step/Ace-Step1.5
	tags:
	- gguf
	- text-to-music
	- acestep
	---

	# acestep — runtime + GGUF weights mirror

	Files needed by the local music generation feature of the KakaoWork desktop app.
	This repository is a mirror kept for build reproducibility: the app pins a
	sha256 for every file, so it needs a location whose bytes never change.

	## Contents

	### Runtime (built by us)

	`acestep.cpp` has no upstream release artifacts, so we build it ourselves.

	\| File \| Platform \|
	\| --- \| --- \|
	\| `acestep-darwin-arm64.zip` \| macOS arm64 (ad-hoc signed) \|
	\| `acestep-win-x64.zip` \| Windows x64 (signed) \|

	Each archive is flat: `ace-lm`, `ace-synth` and the ggml shared libraries sit at
	the top level. The executables resolve the libraries next to them, so keep them
	in one directory.

	### Weights (mirrored, byte-identical)

	Re-uploaded unchanged from
	[Serveurperso/ACE-Step-1.5-GGUF](https://huggingface.co/Serveurperso/ACE-Step-1.5-GGUF).
	Only the four files our app loads are mirrored; that repository has many more
	quantizations and is the place to look for them.

	\| File \| Size \| sha256 \|
	\| --- \| --- \| --- \|
	\| `acestep-v15-turbo-Q4_K_M.gguf` \| 1.45 GB \| `55b4d8514850f3d0f82536f37e99673aaf48df802b5ae5b153eea32a2e2daa5e` \|
	\| `acestep-5Hz-lm-4B-Q5_K_M.gguf` \| 3.03 GB \| `938ed7067c8897f66acf4c3a86fc1fa8113d5cd1a5f13e6edec2e03207514e2d` \|
	\| `Qwen3-Embedding-0.6B-Q8_0.gguf` \| 784 MB \| `972f23255e46adfe744a0eb9a0039f3c63988f65753b0968d776e8b27168c321` \|
	\| `vae-BF16.gguf` \| 337 MB \| `0599862ac5d15cd308e1d2e368373aea6c02e25ebd1737ad4a4562a0901b0ef8` \|

	The sha256 values are the upstream ones — verify against the source repository
	if you want to confirm nothing was altered in transit.

	## Credits and license

	- Original model weights: ACE Studio and StepFun —
	[ACE-Step/Ace-Step1.5](https://huggingface.co/ACE-Step/Ace-Step1.5). All model
	weights are theirs; we changed nothing.
	- GGUF conversion and quantization: Serveurperso —
	[Serveurperso/ACE-Step-1.5-GGUF](https://huggingface.co/Serveurperso/ACE-Step-1.5-GGUF)
	(MIT), whose model card also credits Aryanne and jeankassio for specific
	weight merges.
	- The mirrored weights are redistributed under the MIT license of the source
	repository. Please refer to the upstream repositories for the terms that apply
	to the original model.
	- The runtime archives are builds of
	[acestep.cpp](https://github.com/ACE-Step/acestep.cpp); its own license applies.