okamototk/ gemma-2-llama-swallow

242 Downloads Updated 6 months ago

Gemma 2 Llama Swallow with tool calling(MCP/A2A) support

tools 2b 9b 27b

ollama run okamototk/gemma-2-llama-swallow:2b

curl http://localhost:11434/api/chat \
  -d '{
    "model": "okamototk/gemma-2-llama-swallow:2b",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='okamototk/gemma-2-llama-swallow:2b',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'okamototk/gemma-2-llama-swallow:2b',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Applications

Claude Code

Claude Code ollama launch claude --model okamototk/gemma-2-llama-swallow:2b

Codex

Codex ollama launch codex --model okamototk/gemma-2-llama-swallow:2b

OpenCode

OpenCode ollama launch opencode --model okamototk/gemma-2-llama-swallow:2b

OpenClaw

OpenClaw ollama launch openclaw --model okamototk/gemma-2-llama-swallow:2b

Models

Name

4 models

Size

Context

Input

gemma-2-llama-swallow:2b

1.7GB · 8K context window · Text · 6 months ago

gemma-2-llama-swallow:2b

1.7GB

8K

Text

gemma-2-llama-swallow:9b

5.8GB · 8K context window · Text · 6 months ago

gemma-2-llama-swallow:9b

5.8GB

8K

Text

gemma-2-llama-swallow:27b

17GB · 8K context window · Text · 6 months ago

gemma-2-llama-swallow:27b

17GB

8K

Text

Readme

Gemma 2 Llama Swallow Quantized Model with Tool Calling Support

Gemma 2 Llama Swallow is originally developed by Institute of Science Tokyo and The National Institute of Advanced Industrial Science and Technology.

This Ollama model is quantized as Q4_K_M with imatrix using TFMC/imatrix-dataset-for-japanese-llm.

Add tool calling function for MCP or A2A support for using Agent purpose.

Model(Here)	Original Model	Quantized Model (HF)
okamototk/gemma-2-llama-swallow:2b	Gemma-2-Llama-Swallow-9b-it-v0.1	okamototk/Gemma-2-Llama-Swallow-9b-it-v0.1-gguf
okamototk/gemma-2-llama-swallow:9b	Gemma-2-Llama-Swallow-9b-it-v0.1	okamototk/Gemma-2-Llama-Swallow-9b-it-v0.1-gguf
okamototk/gemma-2-llama-swallow:27b	Gemma-2-Llama-Swallow-27b-it-v0.1	okamototk/Gemma-2-Llama-Swallow-27b-it-v0.1-gguf

License

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

This model is built with Gemma and built with Llama.

# Gemma 2 Llama Swallow Quantized Model with Tool Calling Support

Gemma 2 Llama Swallow is originally developed by Institute of Science Tokyo and The National Institute of Advanced Industrial Science and Technology.

This Ollama model is quantized as Q4_K_M with imatrix using  [TFMC/imatrix-dataset-for-japanese-llm](https://huggingface.co/datasets/TFMC/imatrix-dataset-for-japanese-llm).

Add tool calling function for MCP or A2A support for using Agent purpose.

| Model(Here)               | Original Model                  | Quantized Model (HF) | 
|----------------------|--------------------------|-----------------------|
| okamototk/gemma-2-llama-swallow:2b | [Gemma-2-Llama-Swallow-9b-it-v0.1 ](https://huggingface.co/tokyotech-llm/Gemma-2-Llama-Swallow-2b-it-v0.1) | [okamototk/Gemma-2-Llama-Swallow-9b-it-v0.1-gguf](https://huggingface.co/okamototk/Gemma-2-Llama-Swallow-2b-it-v0.1-gguf)| 
| okamototk/gemma-2-llama-swallow:9b | [Gemma-2-Llama-Swallow-9b-it-v0.1 ](https://huggingface.co/tokyotech-llm/Gemma-2-Llama-Swallow-9b-it-v0.1) | [okamototk/Gemma-2-Llama-Swallow-9b-it-v0.1-gguf](https://huggingface.co/okamototk/Gemma-2-Llama-Swallow-9b-it-v0.1-gguf)| 
| okamototk/gemma-2-llama-swallow:27b | [Gemma-2-Llama-Swallow-27b-it-v0.1 ](https://huggingface.co/tokyotech-llm/Gemma-2-Llama-Swallow-27b-it-v0.1) |[okamototk/Gemma-2-Llama-Swallow-27b-it-v0.1-gguf](https://huggingface.co/okamototk/Gemma-2-Llama-Swallow-27b-it-v0.1-gguf)|

# License
This model is licensed under "LLAMA 3.3 COMMUNITY LICENSE AGREEMENT" and "Gemma Terms of Use". Please check license metadata in model for license detail.

This model is built with Gemma and built with Llama.

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