PeachDB

PeachDB - the AI-First, Embeddings Database

Build memory for your AI products in minutes!

Our core API has 4 functions

from peachdb import PeachDB

# Create a new PeachDB instance or reference an existing one
db = PeachDB(
    project_name="my_app",
    embedding_generator="imagebind",  # "imagebind" or "sentence_transformer_L12"
    embedding_backend="exact_cpu",  # "exact_cpu", "exact_gpu", or "approx"
    distance_metric="cosine",  # "cosine" or "l2"
)

# Auto-compute & upsert embeddings at scale using the specified `embedding_generator` model on Modal
db.upsert_text(  # or "upsert_image" or "upsert_audio"
    csv_path="/path/to/local/csv",  # or "s3://path/to/csv"
    column_to_embed="foo",  # column values can either be string or public URI to image/audio
    id_column_name="id",
    embeddings_output_s3_bucket_uri=None,  # required when using S3 URI for `csv_path`
    max_rows=None,  # or N to process N rows
)

# Query top 5 similar results
ids, distances, results_df = db.query(
    query_input='An example query',  # or path to an image/audio file
    modality='text',  # "text", "image", or "audio"
    top_k=5
)

# Deploy database as a publicly accessible FastAPI server
# GET /query?query_input='An example query'&modality=[text|image|audio]&top_k=5 to fetch 5 most similar results
db.deploy()

Why another embedding database?

We've streamlined the entire end-to-end process of creating, storing, and retrieving embeddings, making it developer-friendly, seamless, and cost-effective. You no longer have to build custom pipelines or fret over hardware setups & scalability issues. PeachDB ensures you can get started within minutes, leaving the worries of cost optimization and scale to us.

Our key features include:

• Automated, cost-effective & large-scale embedding computation: We leverage serverless GPU functions (through Modal) to compute embeddings on a large scale efficiently and affordably.
  • For instance, we processed the Kaggle 5M song lyrics dataset in just 12 minutes at a cost of $4.90, using sentence transformers!
  • We've developed Modal wrappers for:
    • Sentence Transformer L12
    • ImageBind,
    • Coming soon, support for:
      • Open-source embedding models such as OpenCLIP, Microsoft E5-v2, and more.
      • Multi-threaded OpenAI embedding calculation.
  • Coming soon: Bring your own embeddings.
  • Coming soon: Custom embedding functions for even more flexibility.
• Multimodality: Native support for data with mixture of modalities (such as image/audio/video).
• Highly Customizable: PeachDB allows you to tailor its features to your needs. You can customize:
  • Embedding computation: as described above.
  • Backend: choose between exact_cpu (numpy), exact_gpu (torch), or approx (HNSW).
  • Distance metrics: cosine or l2.
• Effortless Deployment: Deploy PeachDB as a publicly available server with a single API. No need to worry about nginx or SSL certs.
  • Coming soon: Managed, scalable deployment.
• Consistent API: Experience the same API across all environments - dev, test, and prod.
• Open Source: Apache 2.0.

Example

Below is a walkthrough of creating a web server for a music recommendation app. To power the app, we are using the Kaggle 5M song lyric dataset

• Ssh into your remote instance (doesn't need GPU)

• Create & activate a new conda environment conda create -n spoti_vibe python=3.10 & conda activate spoti_vibe

• Install PeachDB: pip install peachdb

• Setup Modal

  • Create an account at modal.com
  • Install the modal-client package: pip install modal-client
  • Setup token: modal token new

• (optional: for AWS S3) PeachDB accepts local & S3 paths to datasets for embedding computation. To use S3 URIs, ensure you've installed the aws cli and run aws configure. The credentials should have read & write access to the relevant bucket you plan to use.

• mkdir spoti_vibe/

• Create a new module inside the directory server.py

• Add the following code

  from peachdb import PeachDB
  
  import os
  
  # Fetch the username & key by creating a new API token at https://www.kaggle.com/settings
  os.environ["KAGGLE_USERNAME"] = None  # set user name
  os.environ["KAGGLE_KEY"] = None  # set key
  
  import kaggle  # make sure you've run `pip install kaggle`
  
  kaggle.api.authenticate()
  # It can take a few mins to download depending on the network speed
  kaggle.api.dataset_download_files("nikhilnayak123/5-million-song-lyrics-dataset", path=".", unzip=True)
  
  db = PeachDB(
      project_name="spoti_vibe",
      distance_metric="cosine",
      embedding_backend="exact_cpu",
      embedding_generator="sentence_transformer_L12",
  )
  db.upsert_text(
      csv_path="./ds2.csv",  # dataset name as observed on Kaggle
      column_to_embed="lyrics",
      id_column_name="id",
  )
  
  db.deploy()  # Public URL will be printed to console
  

And that's it! You should now have a publicly available server that can listen to query requests from the user on:
GET <PUBLIC_URL>/query?text='Happy, upbeat summer'&top_k=5

Use-cases

• Build web apps like - clip.audio & awesome-movies.life
• Build ChatGPT for X
• Build ChatGPT plugins

FAQs

Q) How can I delete a project? Run db.delete(project_name="my_app")

Get Involved

We welcome PR contributors and ideas for how to improve the project.

License

Apache 2.0