Llama Swallow with additional Tool Calling support

tools 8b 70b

ollama run okamototk/llama-swallow:8b

curl http://localhost:11434/api/chat \
  -d '{
    "model": "okamototk/llama-swallow:8b",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='okamototk/llama-swallow:8b',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'okamototk/llama-swallow:8b',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Applications

Claude Code ollama launch claude --model okamototk/llama-swallow:8b

Codex ollama launch codex --model okamototk/llama-swallow:8b

OpenCode ollama launch opencode --model okamototk/llama-swallow:8b

OpenClaw ollama launch openclaw --model okamototk/llama-swallow:8b

Models

View all →

Name

2 models

Size

Context

Input

llama-swallow:8b

4.9GB · 128K context window · Text · 6 months ago

llama-swallow:8b

4.9GB

128K

Text

llama-swallow:70b

43GB · 128K context window · Text · 6 months ago

llama-swallow:70b

43GB

128K

Text

Readme

Llama Swallow Tool Calling Supported Quantized Model for Ollama

After MCP was hot, Tool Calling feature is important to develop AI Agent with various external tools. This Ollama model is added Tool Calling capability for Llama Swallow provided by Institute of Science Tokyo.

This model set is Q4_K_M imatrix quantized model. TFMC/imatrix-dataset-for-japanese-llm is used for imatrix calibration.

Model(Here)	Original Model	Quantized Model (HF)
llama-swallow:8b	Llama-3.1-Swallow-8B-Instruct-v0.5	Llama-3.1-Swallow-8B-Instruct-v0.5-gguf
llama-swallow:70b	Llama-3.3-Swallow-70B-Instruct-v0.4	Llama-3.3-Swallow-70B-Instruct-v0.4-imatrix-gguf

License

This model is licensed under “LLAMA 3.1 / 3.3 COMMUNITY LICENSE AGREEMENT” and “Gemma Terms of Use”. Please check license metadata in model for license detail.

This model is Built with Llama.

# Llama Swallow Tool Calling Supported Quantized Model for Ollama

After MCP was hot, Tool Calling feature is important to develop AI Agent with various external tools. This Ollama model is added **Tool Calling capability** for Llama Swallow provided by Institute of Science Tokyo.

This model set is Q4_K_M imatrix quantized model. [TFMC/imatrix-dataset-for-japanese-llm](https://huggingface.co/datasets/TFMC/imatrix-dataset-for-japanese-llm) is used for imatrix calibration.

| Model(Here)               | Original Model                  | Quantized Model (HF) | 
|----------------------|--------------------------|-----------------------|
| llama-swallow:8b | [Llama-3.1-Swallow-8B-Instruct-v0.5 ](https://huggingface.co/tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.5) | [Llama-3.1-Swallow-8B-Instruct-v0.5-gguf](https://huggingface.co/okamototk/Llama-3.1-Swallow-8B-Instruct-v0.5-gguf)| 
| llama-swallow:70b | [Llama-3.3-Swallow-70B-Instruct-v0.4](https://huggingface.co/tokyotech-llm/Llama-3.3-Swallow-70B-Instruct-v0.4) | [Llama-3.3-Swallow-70B-Instruct-v0.4-imatrix-gguf](https://huggingface.co/okamototk/Llama-3.3-Swallow-70B-Instruct-v0.4-imatrix-gguf)|
# License
This model is licensed under "LLAMA 3.1 / 3.3 COMMUNITY LICENSE AGREEMENT" and "Gemma Terms of Use". Please check license metadata in model for license detail.

This model is Built with Llama.

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)