Easy to use CLI to set up embedding-explorer for deployment on a corpus with GloVe word embeddings
Project description
glove-semantic-explorer
Set up embedding explorer on a corpus with GloVe word embeddings with an easy-to-use CLI.
Installation
You can install the CLI from PyPI:
We recommend that you use a Linux/Unix system, preferably Debian when using this tool. Windows and MacOS could still work, but we do not guarrantee this.
pip install glove-semantic-explorer
Usage
1. Train a model
You will need a corpus in the format of a bunch of .txt
files in a folder.
Every line in a file should represent one sentence/passage.
To train a GloVe model on the corpus, run:
python3 -m glove_semantic_explorer train_model dat/ -o model/glove.kv
This will output a keyed vectors file to model/glove.kv
.
2. Run the Explorer
To run the explorer on the trained model locally, run:
python3 -m glove_semantic_explorer run_explorer -m model/glove.kv --port 8080
This will start embedding-explorer on the trained embedding model on port 8080.
3. Deploy!
You can deploy the application using docker compose
.
The way this can be done with our CLI is by auto-generating a Dockerfile
, a compose.yaml
and a main.py
file, that contains all the code for running the server.
To output this into a folder called deployment/
, run the following command:
python3 -m glove_semantic_explorer generate_docker "your_project_name" -m model/glove.kv -p 8080 -o deployment/
Beware that the model file only gets mounted to the container, and thus should not be removed, moved or renamed.
To deploy the app with docker compose run the following:
cd deployment/
sudo docker compose up
The app will then run on port 8080.
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
Hashes for glove_semantic_explorer-0.1.1.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | b3dfb7b3ff021af6ee146cd15a32761a025d011e76fae570cb41fd2ea0fd9ca8 |
|
MD5 | dd574119f75888d5b852e317218af69f |
|
BLAKE2b-256 | 4e22a20834f8fc8dc5400302468abe2615eed103ec9e12f87b139ca881d5afac |
Hashes for glove_semantic_explorer-0.1.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | f36912414d8423e453050cfffecd159b40b3ac08a1193cc4418ffdddd474e23a |
|
MD5 | 77721cda3f9d03abcb5fb2410d14a35f |
|
BLAKE2b-256 | a0484e0b9703ae58370f0be1fc10aaf60a5019e0f2ccbe300e053707954c3bee |