Instructions to use team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit with Transformers:

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

pipe = pipeline("text-generation", model="team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit

SGLang

How to use team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit 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 "team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit" \
    --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": "team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit",
		"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 "team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit" \
        --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": "team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit with Docker Model Runner:
```
docker model run hf.co/team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit
```

Tanuki-8x8B-dpo-v1.0-GPTQ-8bit

概要

GENIAC 松尾研 LLM開発プロジェクトで開発されたLLMであるweblab-GENIAC/Tanuki-8x8B-dpo-v1.0のGPTQ 8bit量子化モデルです。
キャリブレーションセットにはizumi-lab/wikipedia-ja-20230720からランダムサンプリングした1000件を利用しています。

情報提供

この元モデルは独自アーキテクチャを持つモデルであるため、変換のためにAutoGPTQライブラリの一部改変が必要となりました。本モデルの変換に必要な改変を施したAutoGPTQをこちらで公開しています。

また、同様の理由で本モデルをvLLMを使って推論する際にはvLLMライブラリの一部改変が必要となります。本モデルの推論に必要な改変を施したvLLMをこちらで公開しています。
これを以下のようにソースからビルドし推論に利用してください。

git clone https://github.com/team-hatakeyama-phase2/vllm.git
cd vllm
LD_LIBRARY_PATH="" MAX_JOBS=16 pip install -e .

モデルに関する詳細については元モデルをご確認ください。

Downloads last month: 41

Model tree for team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit

Base model

weblab-GENIAC/Tanuki-8x8B-dpo-v1.0

Quantized

(4)

this model

team-hatakeyama-phase2
/

Tanuki-8x8B-dpo-v1.0-GPTQ-8bit

Tanuki-8x8B-dpo-v1.0-GPTQ-8bit

概要

情報提供

Model tree for team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit

Dataset used to train team-hatakeyama-phase2/Tanuki-8x8B-dpo-v1.0-GPTQ-8bit