https://github.com/pinecone-io/basic-operations-workshop
https://github.com/pinecone-io/basic-operations-workshop
Last synced: 4 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/pinecone-io/basic-operations-workshop
- Owner: pinecone-io
- Created: 2023-08-01T00:56:33.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2024-12-23T22:39:16.000Z (6 months ago)
- Last Synced: 2025-01-11T05:21:51.193Z (6 months ago)
- Language: Jupyter Notebook
- Size: 15.9 MB
- Stars: 2
- Watchers: 5
- Forks: 0
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Pinecone Basic Operations Workshop
This workshop contains a collection of labs that will get you hands-on with Pinecone for basic operations. If you would like more use case specific material
please refer to [Pinecone Examples](https://docs.pinecone.io/page/examples). If you are interested in deep dives, refer to [Pinecone Learning Center](https://www.pinecone.io/learn/)
## [Lab #1](./lab1.ipynb)
This is a simple lab that has the basics for setting environment variables, installing dependencies and working with the pinecone client. This lab uses dummy data. This lab introduces the following operations concepts:
* Install the python pinecone grpc client
* Leverage environment variables for pinecone client initialization
* Index creation using a storage optimzed pod(s1)
* Generate and upsert vectors with meta-data
## [Lab #2](./lab2.ipynb)
This lab introduces the following operations concepts:
* Index creation using the x2 pod size (a way to vertically scale your index)
* Use meta-data statistics to learn more about the index
* Index backup/restore using Collections
* Meta-data filter exclusions to guard against high cardinality
This lab also uses dummy data.
## [Lab #3](./lab3.ipynb)
This lab introduces the following operations concepts:
* Load public datasets from Hugging Face
* Generate embeddings using an open source model(CLIP)
* Use training data labels as meta-data for your training data embeddings
* Query pinecone with a test image that is not included in the public data set
* Validate pinecone accuracy with test images that are included in the public data set
* Run a load test using public test data to validate accuracy, P50-P100 latency and QPS
## Troubleshooting
If you encounter issues with python/pip the following commands might help clean up a broken dependency tree.
```
!pip freeze > requirements.txt
!cat requirements.txt | xargs -n 1 pip uninstall -y
```