Instructions to use second-state/Seed-OSS-36B-Instruct-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use second-state/Seed-OSS-36B-Instruct-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="second-state/Seed-OSS-36B-Instruct-GGUF")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("second-state/Seed-OSS-36B-Instruct-GGUF")
model = AutoModelForCausalLM.from_pretrained("second-state/Seed-OSS-36B-Instruct-GGUF")

llama-cpp-python

How to use second-state/Seed-OSS-36B-Instruct-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="second-state/Seed-OSS-36B-Instruct-GGUF",
	filename="Seed-OSS-36B-Instruct-Q2_K.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use second-state/Seed-OSS-36B-Instruct-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf second-state/Seed-OSS-36B-Instruct-GGUF: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 second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf second-state/Seed-OSS-36B-Instruct-GGUF: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 second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M

Use Docker

docker model run hf.co/second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use second-state/Seed-OSS-36B-Instruct-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "second-state/Seed-OSS-36B-Instruct-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "second-state/Seed-OSS-36B-Instruct-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M

SGLang

How to use second-state/Seed-OSS-36B-Instruct-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "second-state/Seed-OSS-36B-Instruct-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "second-state/Seed-OSS-36B-Instruct-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "second-state/Seed-OSS-36B-Instruct-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "second-state/Seed-OSS-36B-Instruct-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use second-state/Seed-OSS-36B-Instruct-GGUF with Ollama:
```
ollama run hf.co/second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
```

Unsloth Studio new

How to use second-state/Seed-OSS-36B-Instruct-GGUF 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 second-state/Seed-OSS-36B-Instruct-GGUF 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 second-state/Seed-OSS-36B-Instruct-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for second-state/Seed-OSS-36B-Instruct-GGUF to start chatting

Pi new

How to use second-state/Seed-OSS-36B-Instruct-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use second-state/Seed-OSS-36B-Instruct-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf second-state/Seed-OSS-36B-Instruct-GGUF: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 second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M

Run Hermes

hermes

Docker Model Runner
How to use second-state/Seed-OSS-36B-Instruct-GGUF with Docker Model Runner:
```
docker model run hf.co/second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
```

Lemonade

How to use second-state/Seed-OSS-36B-Instruct-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull second-state/Seed-OSS-36B-Instruct-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Seed-OSS-36B-Instruct-GGUF-Q4_K_M

List all available models

lemonade list

Seed-OSS-36B-Instruct-GGUF

Original Model

ByteDance-Seed/Seed-OSS-36B-Instruct

Run with LlamaEdge

LlamaEdge version: coming soon

Prompt template

Prompt type:
- seed-oss-think for think mode
- seed-oss-no-think for no think mode

Prompt string

Thinking mode

<seed:bos>system
You are Doubao, a helpful AI assistant.
<seed:eos>

<seed:bos>user
{user_message_1}
<seed:eos>

<seed:bos>assistant
<seed:think>{thinking_content}</seed:think>
{assistant_message_1}
<seed:eos>

<seed:bos>user
{user_message_2}
<seed:eos>

<seed:bos>assistant

No-thinking mode

<seed:bos>system
You are Doubao, a helpful AI assistant.
<seed:eos>

<seed:bos>system
You are an intelligent assistant that can answer questions in one step without the need for reasoning and thinking, that is, your thinking budget is 0. Next, please skip       the thinking process and directly start answering the user's questions.
<seed:eos>

<seed:bos>user
{user_message_1}
<seed:eos>

<seed:bos>assistant
{assistant_message_1}
<seed:eos>

<seed:bos>user
{user_message_2}
<seed:eos>

<seed:bos>assistant

Context size: 512000

Run as LlamaEdge service

wasmedge --dir .:. \
  --nn-preload default:GGML:AUTO:Seed-OSS-36B-Instruct-Q5_K_M.gguf \
  llama-api-server.wasm \
  --prompt-template seed-oss-no-think \
  --ctx-size 512000 \
  --model-name seed-oss

Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
Seed-OSS-36B-Instruct-Q2_K.gguf	Q2_K	2	13.6 GB	smallest, significant quality loss - not recommended for most purposes
Seed-OSS-36B-Instruct-Q3_K_L.gguf	Q3_K_L	3	19.1 GB	small, substantial quality loss
Seed-OSS-36B-Instruct-Q3_K_M.gguf	Q3_K_M	3	17.6 GB	very small, high quality loss
Seed-OSS-36B-Instruct-Q3_K_S.gguf	Q3_K_S	3	15.9 GB	very small, high quality loss
Seed-OSS-36B-Instruct-Q4_0.gguf	Q4_0	4	20.6 GB	legacy; small, very high quality loss - prefer using Q3_K_M
Seed-OSS-36B-Instruct-Q4_K_M.gguf	Q4_K_M	4	21.8 GB	medium, balanced quality - recommended
Seed-OSS-36B-Instruct-Q4_K_S.gguf	Q4_K_S	4	20.7 GB	small, greater quality loss
Seed-OSS-36B-Instruct-Q5_0.gguf	Q5_0	5	25.0 GB	legacy; medium, balanced quality - prefer using Q4_K_M
Seed-OSS-36B-Instruct-Q5_K_M.gguf	Q5_K_M	5	25.6 GB	large, very low quality loss - recommended
Seed-OSS-36B-Instruct-Q5_K_S.gguf	Q5_K_S	5	25.0 GB	large, low quality loss - recommended
Seed-OSS-36B-Instruct-Q6_K.gguf	Q6_K	6	29.7 GB	very large, extremely low quality loss
Seed-OSS-36B-Instruct-Q8_0.gguf	Q8_0	8	38.4 GB	very large, extremely low quality loss - not recommended
Seed-OSS-36B-Instruct-f16-00001-of-00003.gguf	f16	16	30.0 GB
Seed-OSS-36B-Instruct-f16-00002-of-00003.gguf	f16	16	30.0 GB
Seed-OSS-36B-Instruct-f16-00003-of-00003.gguf	f16	16	12.4 GB

Quantized with llama.cpp b6301.

Downloads last month: 128

GGUF

Model size

36B params

Architecture

seed_oss

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

16-bit

Model tree for second-state/Seed-OSS-36B-Instruct-GGUF

Base model

ByteDance-Seed/Seed-OSS-36B-Instruct

Quantized

(39)

this model