Instructions to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- llama-cpp-python
How to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf with llama-cpp-python:
# !pip install llama-cpp-python from llama_cpp import Llama llm = Llama.from_pretrained( repo_id="koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf", filename="sakana-merged-fp16.gguf", )
llm.create_chat_completion( messages = "No input example has been defined for this model task." )
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- llama.cpp
How to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf with llama.cpp:
Install from brew
brew install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf # Run inference directly in the terminal: llama-cli -hf koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf # Run inference directly in the terminal: llama-cli -hf koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
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 koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf # Run inference directly in the terminal: ./llama-cli -hf koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
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 koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf # Run inference directly in the terminal: ./build/bin/llama-cli -hf koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
Use Docker
docker model run hf.co/koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
- LM Studio
- Jan
- Ollama
How to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf with Ollama:
ollama run hf.co/koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
- Unsloth Studio
How to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.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 koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.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 koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf to start chatting
- Pi
How to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama-server -hf koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
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": "koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent new
How to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.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 koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
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 koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
Run Hermes
hermes
- Docker Model Runner
How to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf with Docker Model Runner:
docker model run hf.co/koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
- Lemonade
How to use koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull koguma-ai/SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf
Run and chat with the model
lemonade run user.SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf-{{QUANT_TAG}}List all available models
lemonade list
SakanaAI-TinySwallow-1.5B-Instruct-GRPO
概要
SakanaAI-TinySwallow-1.5B-Instruct-GRPOは、軽量ながら高性能な1.5Bパラメータの指示用言語モデルです。GGUFフォーマットで提供され、効率的な推論が可能です。unslothのGRPO(Gradient-based Reward Policy Optimization)を使用して、モデルの性能を最適化しています。
モデル詳細
- モデル名: SakanaAI-TinySwallow-1.5B-Instruct-GRPO
- バージョン: 1.5B
- タイプ: Instruct(指示用)
- フォーマット: GGUF
- ライセンス: Apache-2.0
- 最適化手法: unsloth GRPO
特徴
- 軽量な1.5Bパラメータ設計
- 指示用(Instruct)モデルとして最適化
- GGUFフォーマットによる効率的な推論
- 日本語と英語の両方に対応
- unsloth GRPOによる高度な最適化
- 勾配ベースの報酬ポリシー最適化
- より自然な応答生成
- 指示への忠実な従順性
使用方法
- 必要なライブラリのインストール
pip install llama-cpp-python
- モデルの読み込みと推論
from llama_cpp import Llama
# モデルの読み込み
llm = Llama(
model_path="SakanaAI-TinySwallow-1.5B-Instruct-GRPO.gguf",
n_ctx=2048, # コンテキストウィンドウサイズ
n_threads=4 # スレッド数
)
# 推論の実行
response = llm(
"こんにちは。今日の天気について教えてください。",
max_tokens=100,
stop=["。", "\n"],
echo=False
)
ライセンス
このモデルはApache-2.0ライセンスの下で提供されています。
謝辞
このモデルの開発に貢献された全ての開発者と研究者に感謝いたします。
- Downloads last month
- 3
We're not able to determine the quantization variants.