iagon-py
is a Python package for storing/accessing data on Iagon, a privacy focused
decentralized storage protocol on Cardano.
Table of Contents
Python is where a significant amount of data science is being done. Especially for artificial intelligence and machine learning. Storage is a necessary component to data driven decentralized applications, so making the first decentralized storage protocol easily accessible to Python opens up a diverse set of integrations.
The biggest driver of this tool right now is storing/loading deep neural networks and
the data needed to train them. Python has a storage abstraction package (fsspec
) that
a number of Python tools use, including PyTorch (an AI framework) and a large body of
data ingestion and transformation tools. The current state of the package is that enough
of the direct API calls to Iagon are implemented to be able to create a virtual file
system for storing/loading PyTorch models and the data used to train them.
Once the virtual file system is created, the rest of the Iagon API will be implemented followed by a more complete implementation of the virtual file system.
You can install with pip
from PyPI, but there is currently a bug in pycardano
that
doesn't allow CIP8
message signing with extended payment keys, so this will not work.
Instead, you will need to clone and install this repo:
pip install iagon-py
A common use case will be sharing public data directly from Iagon. iagon-py
makes it
easy to pull in this data. We will use some clustering data generated for the Minswap
Catalyst Proposal Selection group for Catalyst Fund 10.
The Cardano Catalyst Fund 10 voting on Minswap used OpenAI and some custom clustering methods to aggregate proposals into similar groups to summarize and filter them. This data has been stored on Iagon, so this example will download them all. Some information on how the groups were generated and clustered can be found on Twitter.
Next, create a short script to pull in and store the compressed data.
from pycardano import HDWallet
from iagon import IagonAdapter
# Give the file id of the Minswap Catalyst Grouping data
file_id = "65296cf4eba1933b118b368f"
# Create a random seed phrase to access Iagon
# Why is this needed? To access Iagon, we need to get an auth token using CIP8, which
# requires a wallet signature. Since it's public data, it doesn't matter what wallet we
# use, so generate a random one.
seed_phrase = HDWallet.generate_mnemonic()
# Create the Iagon session
with IagonAdapter.session(seed_phrase) as session:
# Open a file to write the binary data to
with open("minswap_catalyst_groups.zip", "wb") as fw:
fw.write(session.download(file_id=file_id))
Now you can unzip the text files and see what the Minswap Catalyst Selection Group used.
The text files are randomly assigned a number indicating the cluster group number, with
a special group (-1
) that acted as a catch all for proposals that didn't seem to fit
into any group.
A good example of how to upload data will come, but it is currently possible to upload data, create directories and subdirectories, and browse through data on Iagon.
For now, take a look at the code used to upload the Minswap Catalyst Selection Group
zip file: examples\minswap_fund10_upload.py
.
Note that the code requires creating an .env
file with a SEED
value that corresponds
to the wallet you want to use to upload the data.
I can always use volunteers to take on specific chunks of the project. I work on this in my free time, along with some other Cardano projects. You can help by reaching out on Twitter or Discord. Alternatively, sending tips is also helpful to cover the costs of production. Tips can be sent to:
addr1q9hw8fuex09vr3rqwtn4fzh9qxjlzjzh8aww684ln0rv0cfu3f0de6qkmh7c7yysfz808978wwe6ll30wu8l3cgvgdjqa7egnl