Instructions to use palette-lab/songgot-x-0.8b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use palette-lab/songgot-x-0.8b 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 palette-lab/songgot-x-0.8b:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf palette-lab/songgot-x-0.8b:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf palette-lab/songgot-x-0.8b:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf palette-lab/songgot-x-0.8b: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 palette-lab/songgot-x-0.8b:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf palette-lab/songgot-x-0.8b: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 palette-lab/songgot-x-0.8b:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf palette-lab/songgot-x-0.8b:Q4_K_M

Use Docker

docker model run hf.co/palette-lab/songgot-x-0.8b:Q4_K_M

LM Studio
Jan

vLLM

How to use palette-lab/songgot-x-0.8b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "palette-lab/songgot-x-0.8b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "palette-lab/songgot-x-0.8b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/palette-lab/songgot-x-0.8b:Q4_K_M

Ollama
How to use palette-lab/songgot-x-0.8b with Ollama:
```
ollama run hf.co/palette-lab/songgot-x-0.8b:Q4_K_M
```
Unsloth Desktop

How to use palette-lab/songgot-x-0.8b with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf palette-lab/songgot-x-0.8b:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "palette-lab/songgot-x-0.8b:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use palette-lab/songgot-x-0.8b with Docker Model Runner:
```
docker model run hf.co/palette-lab/songgot-x-0.8b:Q4_K_M
```

Lemonade

How to use palette-lab/songgot-x-0.8b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull palette-lab/songgot-x-0.8b:Q4_K_M

Run and chat with the model

lemonade run user.songgot-x-0.8b-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use palette-lab/songgot-x-0.8b with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf palette-lab/songgot-x-0.8b:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default palette-lab/songgot-x-0.8b:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use palette-lab/songgot-x-0.8b with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf palette-lab/songgot-x-0.8b:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "palette-lab/songgot-x-0.8b:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Songgot-X 0.8B

Line B of the Songgot project: built on an open base, not from scratch. Songgot-X is Qwen/Qwen3.5-0.8B (Apache 2.0) post-trained by Palette on the Songgot v10 tool-calling set (263,242 rows, one epoch) with the base model's own chat template and tool-call format. The from-scratch Songgot models live at palette-lab/songgot-12l, songgot-m and songgot-l; this repository exists so that a customer who needs the best call accuracy today can have it, with the lineage stated.

Result (v10 weights, measured 2026-09-15; every model scored with the same lenient parser)

FunctionChat-Bench SingleCall (Kakao, 500 Korean items, 5 tool conditions), exact match on function name and arguments under the public scorer in github.com/hanishkeloth/songgot (eval/functionchat_exact.py), not the benchmark's GPT-4 judge. The parser accepts a bare <function=...> block for every model alike.

model	params	exact	4_random	4_close	8_random	8_close	all	name only
Kanana-2-1.3B-Instruct (Kakao)	1.3B	77.0	76.0	71.0	73.0	69.0	73.2	97.2
Songgot-X 0.8B v10 (this model)	0.8B	74.0	68.0	65.0	66.0	64.0	67.4	93.6
EXAONE-4.0-1.2B (LG)	1.28B	73.0	65.0	53.0	66.0	58.0	63.0	85.2
Songgot-X 0.8B v8 (previous weights, 2026-09-13)	0.8B	67.0	66.0	60.0	62.0	54.0	61.8	94.8
Qwen3.5-2B + the v8 rows (not released)	2B	66.0	65.0	61.0	60.0	54.0	61.2	93.8
Qwen3.5-0.8B (the base, zero-shot)	0.8B	51.0	48.0	41.0	52.0	34.0	45.2	73.8
Songgot 12L (from scratch)	50M	44.0	39.0	30.0	35.0	17.0	33.0	73.6

Kakao authored FunctionChat-Bench, so Kanana's figure may reflect in-distribution training. Standard error at these accuracies is about 2.1 points on 500 items. Kanana-2 is released under the Kanana Open License, which does not allow on-device redistribution; Songgot-X is the best Apache-2.0 model in this table.

v10 against the previous weights, item by item: 70 items fixed, 42 lost, 267 unchanged. Of v10's 163 misses, 131 pick the right tool with a wrong argument value (v8: 165), 19 pick a wrong tool (v8: 25), and 13 decline to call although a matching tool is offered (v8: 1). The name-only score fell 1.2 points for that last reason. The remaining argument misses concentrate on relative datetimes, on copying a text payload after an instruction prefix ("...세줘: 문장"), and on one 19xx year rewritten as 2026; those are the targets of the next round.

Run it offline

GGUF files in this repository: songgot-x-q4_k_m.gguf (505 MB) and songgot-x-q8_0.gguf (774 MB), for llama.cpp, Ollama, LM Studio and the browser. Songgot Pocket (hanishkeloth.github.io/songgot/app) runs this model inside the browser on desktop and Android with no server; pick "Songgot-X 0.8B" in the model selector. Palette Desktop, Palette's native offline app, runs it as a local llama-server sidecar. The prompt format is the Qwen3.5 tool template exactly as rendered by the tokenizer's chat template with tools=[...] and thinking off.

Measured in Pocket on 2026-09-15 with these weights (Chrome, Apple M5 Max, single thread, Q4_K_M): 2.0 to 2.2 s per request from prompt to parsed call, 742 to 764 prompt tokens, 28 to 41 generated, 13 to 21 tokens per second (the 2026-09-14 figure with the v8 weights was 3.5 s at 12.2 tokens per second on the same machine). The GGUF header is written without the converter's phantom 25th multi-token-prediction block (the defect that made the 2026-09-13 files unloadable). In browser runtimes keep the batch size small (64): the 248,320-entry vocabulary makes the logits buffer vocab x batch x 4 bytes.

Training data (v10)

175,000 rows sampled from the Songgot v8 set (palette-lab/songgot-tools-ko): Korean tool-calling pairs written by Palette's own open teacher (Palette-K-Midm) from invented tool schemas and verified by a second pass; no closed model produced any label.
31,307 targeted synthetic rows (weight 2) written by the same teacher and checked by deterministic rules, aimed at the four failure classes of the v8 model: copying names verbatim, from/to and sender/receiver roles, Korean number words, two-digit years.
10,861 human Korean voice-assistant requests from Amazon MASSIVE 1.1 ko-KR train split (CC BY 4.0), mapped onto the massive-agents function declarations (weight 2). The MASSIVE dev and test splits were not used.
KoSGD dialogue-state rows.

Single-tool rows receive distractor tools (1, 4 or 8; name-related or random) so the training conditions mirror the benchmark; 30 percent are restyled and 5 percent become negatives. Benchmark tool names and queries are excluded by an assertion at assembly time. Details, the recipe and every negative result are in the Songgot paper (hanishkeloth.github.io/songgot).

Usage

Standard Qwen3.5 chat template with tools=[...]; thinking off. Tool calls come back as <tool_call><function=NAME><parameter=KEY>VALUE</parameter>...</function></tool_call>; booleans and numbers are JSON-typed. When no offered tool fits, the model answers in Korean that no tool matches; stop generation at </tool_call> or at the end of the first line.

Licence and lineage

Apache 2.0, inheriting Qwen3.5-0.8B's Apache 2.0. The base model was trained by Alibaba; Palette trained the post-training stage only. MASSIVE is CC BY 4.0 (Amazon). Answers are generated by an AI model (AI 기본법 labelling).

Downloads last month: 310

Safetensors

Model size

0.8B params

Tensor type

F32

Model tree for palette-lab/songgot-x-0.8b

Base model

Qwen/Qwen3.5-0.8B-Base

Finetuned

Qwen/Qwen3.5-0.8B

Quantized

(260)

this model

palette-lab
/

songgot-x-0.8b

Songgot-X 0.8B

Result (v10 weights, measured 2026-09-15; every model scored with the same lenient parser)

Run it offline

Training data (v10)

Usage

Licence and lineage

Model tree for palette-lab/songgot-x-0.8b

Dataset used to train palette-lab/songgot-x-0.8b