A framework for all things related to embeddings across heterogeneous data.
Project description
Binah is a system for training multi-modal embeddings jointly. We develop an image embedding and a text embedding where objects of similar abstract meaning are near each other in a shared vector space. With this we are able create image and video search using arbitrary language.
Screenshot
Installation
pip install binah
Usage
Binah comes bundled with the etz command line tool.
To completely set up the project, simply run:
etz up
This command sets up the whole Binah system de novo. It takes days to run since it downloads datasets and trains complicated deep learning models.
After “etz up” is finished you can start the image search web demo. Simply run:
etz run search
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
File details
Details for the file binah-0.0.1.tar.gz
.
File metadata
- Download URL: binah-0.0.1.tar.gz
- Upload date:
- Size: 2.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/1.12.1 pkginfo/1.4.2 requests/2.18.4 setuptools/39.2.0 requests-toolbelt/0.8.0 tqdm/4.23.4 CPython/3.6.5
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 79049909eaf5c8f59446fbf58efbefd5ba7acb39047f367f99af3d27b0e193f4 |
|
MD5 | f769fff642d692ad4d3dd98e526aec29 |
|
BLAKE2b-256 | 149a1fb860ef01a23fc425e9368844a430e484b8ce7c49849cb1485c32a026c2 |