loog

GPT-4o mini を試す

July 21, 2024

オムニのミニがでました。

https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence

OpenAI Playground で試しました。

https://platform.openai.com/playground/chat?models=gpt-4o-mini

挨拶

USER: おはよう
ASSISTANT: おはようございます!今日はどんなことをお手伝いできますか?

コーディング

コーディングさせてみます。

USER: BS4 を使用して、cli 引数に url, CSS セレクター を受取り、text を全て返す python code を書いて。

ASSISTANT:
BeautifulSoup4(BS4)を使って、CLI引数としてURLとCSSセレクターを受け取り、その要素のテキストを全て返すPythonコードの例を以下に示します。このコードは、`requests`ライブラリを使用してWebページを取得し、指定したCSSセレクターに基づいてテキストを抽出します。

まず、必要なライブラリをインストールします。以下のコマンドを実行してください。

```bash
pip install requests beautifulsoup4
```

次に、以下のPythonスクリプトを作成します。

```python
import sys
import requests
from bs4 import BeautifulSoup

def fetch_text_from_url(url, css_selector):
    try:
        # URLからページを取得
        response = requests.get(url)
        response.raise_for_status()  # HTTPエラーがあれば例外を投げる

        # BeautifulSoupでHTMLを解析
        soup = BeautifulSoup(response.text, 'html.parser')

        # CSSセレクターに基づいて要素を取得
        elements = soup.select(css_selector)
        
        # 各要素のテキストを結合して返す
        return '\n'.join(element.get_text() for element in elements)
    
    except requests.exceptions.RequestException as e:
        return f"Error fetching the URL: {e}"
    except Exception as e:
        return f"An error occurred: {e}"

if __name__ == '__main__':
    if len(sys.argv) != 3:
        print("Usage: python script.py <url> <css_selector>")
        sys.exit(1)

    url = sys.argv[1]
    css_selector = sys.argv[2]

    text = fetch_text_from_url(url, css_selector)
    print(text)
```

### 使用方法
このスクリプトを `scraper.py` として保存した後、次のようにコマンドラインから実行します。

```bash
python scraper.py <URL> <CSSセレクター>
```

例:
```bash
python scraper.py https://example.com 'h1'
```

このコマンドは指定したURLから`<h1>`タグのテキストを抽出して表示します。

### 注意
- リクエストが成功したかどうかや、CSSセレクターに適合する要素が存在するかどうかを確認するためのエラーハンドリングを含めています。
- 必要に応じて、CSSセレクターやURLを変更してください。

このクオリティで、早くて安い・・・となると、
現状不安定なローカルLLMの立ち位置が、更にゆるぎますね。