csv-to-embeddings-model

Trains a model on top of a sbert's pertained models with given trained pairs to be used with Python's Sentence Transformer

Getting Started

To start, run:

pip install -r requirements.txt

Creating Training Pairs

Make sure to create a pairs.csv file with the trained pairs you want to train the model with.

• Example can be found at example_pairs.csv

Generating a Trained Model

After you have installed the requirements and created a pairs.csv file, run:

python model.py

A new, trained model will be stored under trained_model/.

Using sbert Pretrained Models

You may use other pretrained models from here: https://www.sbert.net/docs/pretrained_models.html

The default model used is multi-qa-MiniLM-L6-cos-v1.