Instructions to use abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged with Transformers:

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

pipe = pipeline("text-generation", model="abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged")
model = AutoModelForCausalLM.from_pretrained("abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged

SGLang

How to use abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-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 "abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-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": "abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-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 "abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-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": "abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged with Docker Model Runner:
```
docker model run hf.co/abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged
```

SO0529 commited on Apr 20, 2024

Commit

fbdcc4d

verified ·

1 Parent(s): 45a9163

Update README.md

Browse files

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -9,7 +9,7 @@ widget:
 # Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged
 Mixtral-8x7B-Instruct-v0.1-japanese-alpha-mergedは[Mixtral-8x7B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1)をベースに日本語の語彙拡張継続事前学習を実施した[学習途中のモデル](https://huggingface.co/abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha)に対して、差分マージを実施したモデルです。
-[ABEJAのテックブログ](https://tech-blog.abeja.asia/entry/abeja-nedo-project-202404)にて評価を実施した途中結果モデルとして公開しています。
 学習を実施したMetagton-LMのレポジトリは[こちら](https://github.com/abeja-inc/Megatron-LM)です。

 # Mixtral-8x7B-Instruct-v0.1-japanese-alpha-merged
 Mixtral-8x7B-Instruct-v0.1-japanese-alpha-mergedは[Mixtral-8x7B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1)をベースに日本語の語彙拡張継続事前学習を実施した[学習途中のモデル](https://huggingface.co/abeja/Mixtral-8x7B-Instruct-v0.1-japanese-alpha)に対して、差分マージを実施したモデルです。
+[ABEJAのテックブログ](https://tech-blog.abeja.asia/entry/abeja-nedo-project-part1-202404)にて評価を実施した途中結果モデルとして公開しています。
 学習を実施したMetagton-LMのレポジトリは[こちら](https://github.com/abeja-inc/Megatron-LM)です。