Skip to content

Latest commit

 

History

History
85 lines (63 loc) · 1.84 KB

README.md

File metadata and controls

85 lines (63 loc) · 1.84 KB
Raw

QuickStart(Python version 3.8.18)

  1. Setup environment variable JWT token from pinata cloud. Download kaggle.json from kaggle.com and place it on your root directory. Make sure to verify you kaggle account.

  2. Create virtual environment

python3 -m venv .venv

  1. Use virtual environment

source .venv/bin/activate

  1. Install dependencies

pip install -r requirements

  1. Start the server

python3 app.py

Local Environment testing.

  1. Enter your file URL and set GPU and TPU as true or false. curl -X GET "http://localhost:3000/compute?fileUrl=<your_file_url>&enable_gpu=<true/false>&enable_tpu=<true/false>" Output: folderID

  2. Enter your folderID to check status of the compute
    curl -X GET "http://localhost:3000/status?folderID=<folderID>" Output: Running/Complete

  3. Enter your folderID to store compute output logs on IPFS CID curl -X GET "http://localhost:3000/output?folderID=<folderID>" Output: CID

  4. Enter your CID on the URL to view the log https://gateway.pinata.cloud/ipfs/<yourCIDhere>

Example(On test API)

  1. Enter your URL replacing sample URL curl -X GET "https://kkpy.onrender.com/compute?fileUrl=https://raw.githubusercontent.com/madhukar123456/kaggle-kernel/main/python-code/mistraltest.ipynb" Output: 3dbbb8fa-e3eb-4948-80c1-8a071e1b55f8

  2. Enter the folder ID to check status curl -X GET "https://kkpy.onrender.com/status?folderID=3dbbb8fa-e3eb-4948-80c1-8a071e1b55f8" Output: Running Output: Complete

  3. After you see output complete in few minutes, enter the folder ID to receive the CID

curl -X GET "https://kkpy.onrender.com/output?folderID=3dbbb8fa-e3eb-4948-80c1-8a071e1b55f8"

Output: QmPDm74QL6kaToB2VEmcTwqRjk53GeqRFVwUhZno2pptQy

  1. You can check your output log placing it on the URL as example below:

https://gateway.pinata.cloud/ipfs/QmNhakdKbJyAVGJ4WmGBxJ3yfnrWAJ1HKk1Liksarzb4Ax