Instructions to use azuki-digital/gemma-2-27b-it-2_lora_merged with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use azuki-digital/gemma-2-27b-it-2_lora_merged with Transformers:

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

pipe = pipeline("text-generation", model="azuki-digital/gemma-2-27b-it-2_lora_merged")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("azuki-digital/gemma-2-27b-it-2_lora_merged")
model = AutoModelForCausalLM.from_pretrained("azuki-digital/gemma-2-27b-it-2_lora_merged", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use azuki-digital/gemma-2-27b-it-2_lora_merged with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "azuki-digital/gemma-2-27b-it-2_lora_merged"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "azuki-digital/gemma-2-27b-it-2_lora_merged",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/azuki-digital/gemma-2-27b-it-2_lora_merged

SGLang

How to use azuki-digital/gemma-2-27b-it-2_lora_merged 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 "azuki-digital/gemma-2-27b-it-2_lora_merged" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "azuki-digital/gemma-2-27b-it-2_lora_merged",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "azuki-digital/gemma-2-27b-it-2_lora_merged" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "azuki-digital/gemma-2-27b-it-2_lora_merged",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Unsloth Desktop
Docker Model Runner
How to use azuki-digital/gemma-2-27b-it-2_lora_merged with Docker Model Runner:
```
docker model run hf.co/azuki-digital/gemma-2-27b-it-2_lora_merged
```

学習データ

以下のデータセットを使用。

llm-jp/magpie-sft-v1.0 (apache-2.0)
DeL-TaiseiOzaki/Tengentoppa-sft-qwen2.5-32b-reasoning-100k (apache-2.0)
weblab-GENIAC/Open-Platypus-Japanese-masked (MIT)
- MITライセンスのデータのみ抽出して使用。

gemma-2利用にあたり、ライセンス上制約の懸念のあるデータセットは利用していない。

google/gemma-2-27bにLoraアダプタ（Taka2024/gemma-2-27b-it-2_lora）をマージ(16bit)して保存したもの。

Downloads last month: 16

Safetensors

Model size

27B params

Tensor type

BF16

Model tree for azuki-digital/gemma-2-27b-it-2_lora_merged

Base model

google/gemma-2-27b

Finetuned

(29)

this model

azuki-digital
/

gemma-2-27b-it-2_lora_merged

学習データ

Model tree for azuki-digital/gemma-2-27b-it-2_lora_merged

Datasets used to train azuki-digital/gemma-2-27b-it-2_lora_merged