Osusume (おすすめ)

Text-Based Recommendation Engine using Cohere

Osusume is a library to implement text-based recommendation engine using Cohere Embeddings and Generation API. Initally was built for a Hacktahon Submission hosted by Lablab.AI and Cohere.

View Changelog

Quick Start

import pandas as pd

import nltk
nltk.download('wordnet')
nltk.download('omw-1.4')

from osusume import Osusume

anime_df = pd.read_json('anime_sm.json')

engine = Osusume.from_dict({
    'token': '<cohere-token>',
    'dataset': anime_df,
    'sentiment_column': 'synopsis',
    'response_prompt': 'anime_data.txt',
    'response_column': [
        ('Title', 'title'),
        ('Synopsis', 'synopsis')
    ]
}).fit()

engine.predict(
    query='animes about monsters and aliens from another world', 
    n_out=3
)

# Top 3 Anime's from given Query

Installation

~ pip install osusume --upgrade

You need to use Pip to install osusume. Conda package is currently unavailable.

Requirements

Python >= 3.8
Cohere
Pandas
Numpy
Scikit Learn
NLTK

Author

Abhishta Gatya (Email) - Software and Machine Learning Engineer

Metadata

Release files for osusume 1.0.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for osusume 1.0.1
File	Size	Uploaded
osusume-1.0.1.tar.gz	12.1 kB	Dec 9, 2022	Details

Built distribution (wheel)

Table of built distributions (wheels) for osusume 1.0.1
File	Interpreter	ABI	Platform	Reset
osusume-1.0.1-py3-none-any.whl 16.9 kB Dec 9, 2022	Python 3	none	any	Details

Total release size: 29.0 kB

Release files / osusume-1.0.1.tar.gz

Download URL	osusume-1.0.1.tar.gz
Size	12.1 kB
Tags	Source
SHA-256 checksum How to use checksums	`9df21335a66f1b23d134fbce66ef2c1c6443a7968e93dbc620016caf8a27d367`
BLAKE2b-256 checksum How to use checksums	`6b482c9354ec06ae57ee16e7fa26e545a07ce7f8ca1304972c7ea76368f4631d`
Upload date	Dec 9, 2022
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/4.0.2 CPython/3.8.9`

Release files / osusume-1.0.1-py3-none-any.whl

Download URL	osusume-1.0.1-py3-none-any.whl
Size	16.9 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`dd5e6827f7827ba71f1f37e7dcc9e8ad5d3b0d0b0fbf5d1b74866c53347c0431`
BLAKE2b-256 checksum How to use checksums	`195976947758c88b54f642e40452a7eb39eac95e710ac20f54096f93b1f045b4`
Upload date	Dec 9, 2022
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/4.0.2 CPython/3.8.9`