Hugging Faceモデルの実行

Hugging Faceを使いこなす

Jacob H. Marquez

Lead Data Engineer

Hugging Faceを使った推論

ローカル環境を使った推論

推論プロバイダー

¹ https://huggingface.co/docs/inference-providers/en/index

Hugging Faceを使った推論

ローカル環境

🆓 無料
😃 手軽
😴 低速、リソースの大量消費

推論プロバイダー

💨高速
🆓 無料で始められる

¹ https://huggingface.co/docs/inference-providers/en/index

Transformersライブラリ

事前学習済みモデルを簡単に利用できる

¹ https://github.com/huggingface/transformers

パイプライン

from transformers import pipeline


gpt2_pipeline = pipeline(task="text-generation", model="openai-community/gpt2")


print(gpt2_pipeline("What if AI"))

[{'generated_text': 'What if AI wouldn\'t be used?"\n\nI had to agree with your
  theory. If a machine\'s learning algorithm is a perfect match for all of a
  human\'s needs, then you may not have a problem with it. My problem was whether'}]

¹ モデルカード: https://huggingface.co/openai-community/gpt2

パイプラインパラメーターの調整

from transformers import pipeline

gpt2_pipeline = pipeline(task="text-generation", model="openai-community/gpt2")


results = gpt2_pipeline("What if AI", max_new_tokens=10, num_return_sequences=2)


for result in results:
    print(result['generated_text'])

What if AI had never existed?  
What if AI could be really smarter than us?

推論プロバイダーの使い方

import os
from huggingface_hub import InferenceClient

client = InferenceClient(
    provider="together",
    api_key=os.environ["HF_TOKEN"],
)

¹ https://huggingface.co/docs/inference-providers/en/index

completion = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V3",
    messages=[
        {
            "role": "user",
            "content": "What is the capital of France?"
        }
    ],
)

¹ https://huggingface.co/docs/inference-providers/en/index

print(completion.choices[0].message)

The capital of France is **Paris**. It is known for its iconic landmarks such as
the Eiffel Tower, the Louvre Museum, and Notre-Dame Cathedral.  

Would you like any additional information about Paris or France?

練習しましょう！

Hugging Faceを使いこなす