Description needed
Project description
CLTrier PromptClassify
Usage
Python Module
from cltrier_promptClassify import Pipeline
# init pipeline object
pipeline = Pipeline({
# pipeline objects
'do_classification': True,
'do_evaluation': True,
'do_export': True,
# path to export dir (only if do_export)
'export_path': './path/dir/',
# dataset configuration
'dataset': {
# path to data file (.csv)
'path': './path/file.csv',
# column containing src text
'text_column': 'text',
# column containing gold label (only if do_evaluation)
'gold_column': 'gold',
# (optional) batch size used during classification
'batch_size': 16,
},
# classifier configuration
'classifiers': [
# label for export, slug/url from hugging face hub
['model_label', 'model_huggingface_slug'],
# ...
],
'templates': [
# {classes}, {text} are dynamically replaced during runtime
['template_label', 'prompt_template (must include {classes} and {text})'],
# ...
],
# list of classes to use
'classes': ['class_1', 'class_2']
})
# call pipeline object
pipeline()
Terminal Script
python3 -m cltrier_promptClassify ./path/to/config.toml
# pipeline objects
do_classification = true
do_evaluation = true
do_export = true
# path to export dir
export_path = './path/dir/'
# dataset configuration
[dataset]
# path to data file (.csv)
path = './path/file.csv'
# column containing src text
text_column = 'text'
# column containing gold label (only if do_evaluation)
gold_column = 'gold'
# (optional) batch size used during classification
batch_size = 16
# classifier configuration
[classify]
# label for export, slug/url from hugging face hub
models = [
['model_label', 'model_huggingface_slug'],
# ...
]
# {classes}, {text} are dynamically replaced during runtime
templates = [
['template_label', 'prompt_template (must include {classes} and {text})'],
]
# list of classes to use
classes = ['class_1', 'class_2']
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Close
Hashes for cltrier_promptclassify-0.0.1.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 9387ee757765d686cddd793edf59d6a6146cf5f053fd8ce332212cf9a7aac7d8 |
|
MD5 | 1920bb24ffce9e0bf05c42702fd893ba |
|
BLAKE2b-256 | a923fd6a22219acbb0ae46a1cd1f5dc0f3c100786b944e33dbbb2bafe2956389 |
Close
Hashes for cltrier_promptclassify-0.0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | cc00c744be4fe5099799652c475371649500f7b2c7162716a3a90e654f658c57 |
|
MD5 | 3e7b9f6ad266fd9c27a0d917db9ccc39 |
|
BLAKE2b-256 | eca8b82bab0e4ad8ce3c166215e59fa4cb579557dad5f7dfc82ebe5e41ec1716 |