Skip to main content

SemanticOwl + LightOwl + LangOwl open-core

Project description

pipowl 🦉

Open Semantic Tools for Python

PyPI version License Downloads

SemanticOwl + LightOwl + LangOwl Open-Core


pipowl 提供:

SemanticOwl

輕量語意編碼器(使用 SentenceTransformer)

LightOwl

文本清洗工具

LangOwl

語意搜尋(top-k + cosine similarity)


安裝

pip install pipowl
使用方式
① 叫出 LangOwl(像叫一隻貓頭鷹出來)
python
複製程式碼
from pipowl.lang import LangOwl

lang = LangOwl() 準備一些要比對的句子清單
python
複製程式碼
corpus = [
    "我今天真的好累",
    "我覺得今天狀態不太好",
    "今天的天氣真的很好",
] 丟一句話進去,看哪句最像
python
複製程式碼
results = lang.topk("我今天真的很想睡覺,因為工作太累了", corpus) 印出結果(分數+句子)
python
複製程式碼
for text, score in results:
    print(score, text)
總結

pipowl 是一個用來「比較句子語意相似度」的小工具。

你只要給它兩個東西:

你的句子(要查的)

你的一組句子(要比的)

它就會告訴你「哪一句最像」。

輸出示例
0.903 我今天真的好累
0.882 我覺得今天狀態不太好
0.834 今天的天氣真的很好

快速試用(Quickstart)
py -m quickstart

quickstart.py
from pipowl.lang import LangOwl

# 叫出語意偵探 LangOwl
lang = LangOwl()

# 準備一組你想比較的句子清單
corpus = [
    "我今天真的好累",
    "我覺得今天狀態不太好",
    "今天的天氣真的很好",
]

print("模型載入完成!你可以開始輸入句子(Ctrl+C 結束)\n")

# 互動式輸入
while True:
    query = input("請輸入句子: ")

    results = lang.topk(query, corpus)

    print("\n相似結果:")
    for text, score in results:
        print(f"{score:.3f} | {text}")
    print()

minimal.py
from pipowl.lang import LangOwl

lang = LangOwl()
results = lang.topk("我很累", ["我好累", "我想吃飯"])
print(results)

```bash

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pipowl-1.4.7.tar.gz (9.3 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

pipowl-1.4.7-py3-none-any.whl (9.1 kB view details)

Uploaded Python 3

File details

Details for the file pipowl-1.4.7.tar.gz.

File metadata

  • Download URL: pipowl-1.4.7.tar.gz
  • Upload date:
  • Size: 9.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.10

File hashes

Hashes for pipowl-1.4.7.tar.gz
Algorithm Hash digest
SHA256 8f33c88579b5f2f861b54cc9b4337d36af85cf384d7d4c712a6c3ba6efbc466f
MD5 0f554fa53c4c870440c69d6f48b954ac
BLAKE2b-256 6ca6d38ffab8806c3156b65ffa582f085742289fa503e45cf9745658560383e7

See more details on using hashes here.

File details

Details for the file pipowl-1.4.7-py3-none-any.whl.

File metadata

  • Download URL: pipowl-1.4.7-py3-none-any.whl
  • Upload date:
  • Size: 9.1 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.10

File hashes

Hashes for pipowl-1.4.7-py3-none-any.whl
Algorithm Hash digest
SHA256 3f1529a60eb8ea82b78f3a605501b7b2da1be3b3b9a16e6cbbb67a35c6613476
MD5 e97e9d0e1d8ae4afa2c7b9f48dec60f5
BLAKE2b-256 3400d8e90629ffcf9222c74d8a4b5f26240e88219ace1cf4ea136d07aa9a666f

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page