cog VS web3.storage

View on GitHub

I'm in the same situation. I found this cog project to dockerise ML https://github.com/replicate/cog : you write just one python class and a yaml file, and it takes care of the "CUDA hell" and deps. It even creates a flask app in front of your model.

That helps keep your system clean, but someone with big $s please rewrite pytorch to golang or rust or even nodejs / typescript.

https://github.com/replicate/cog

Our thinking was just that a bunch of folks will want to fine-tune right away, then deploy the fine-tunes, so trying to make that easy... Or even just deploy the models-as-is on their own infra without dealing with CUDA insanity!

I have used this tool before: https://github.com/replicate/cog/tree/main

Then I discovered Cog: Containers for Machine Learning. Looks like a way more flexible solution to plug in the existing infrastructure: you write your custom code and Cog plugs it in a Docker image with FastAPI, no extra ecosystem complexity added.

Highly recommend using cog https://github.com/replicate/cog

Dokku Maintainer here.

Dokku also supports Dockerfiles, Docker Images, Tarballs (similar to heroku slugs), and Cloud Native Buildpacks. I'm also actively working on AWS Lambda support (both for simple usage without much config as well as SAM-based usage) and investigating Replicate's Cog[1] and Railways Nixpacks[2] functionalities for building apps.

There are quite a few options in the OSS space (as well as Commercial offerings from new startups and popular incumbents). It's an interesting space to be in, and its always fun to see how new offerings innovate on existing solutions.

  [1] https://github.com/replicate/cog

https://web3.storage/ https://estuary.tech/

I'm building Flash — a service to deploy websites and apps on the new decentralized stack. It relies on public infrastructure (such as Estuary, web3.storage and others) instead of providing its own, making the bandwidth and storage very cheap and accessible.

https://web3.storage/ does everything you need

Web3.Storage and other similar projects run a network of IPFS nodes that allow users to store whatever content they like, as a kind of competitor to Dropbox or Google Drive: this could be their personal content, or again, any application content, because some apps integrate directly with this service via their API

You can file this as an issue in the repo here and perhaps mention to them that, as a user, it feels phishy to you.

Typically app developers would be the ones to build on Filecoin and you would choose an app like web3.storage to store your data. As a consumer you dont have to directly hold FIL but you benefit from the Filecoin ecosystem.

Filecoin may not have a front-end but our team has worked on https://estuary.tech and it has been easy for our users that have an invite to onboard to the Filecoin Network, another great option is https://web3.storage

clone repos to your computer, dont just fork them

the fork disappears when the parent repository is taken down, clones stay on your machine

upload to IPFS. you can pin on IPFS for free with filecoin. Use filecoin nodes to pin on ipfs at https://web3.storage

As for storing data on Filecoin directly, that would require making a deal with a Filecoin Storage Provider; doing so is currently a complex task but reading the documentation should get you somewhere. You will need a small amount of the FIL token to make the deal with an SP, who will then store your data. Keep in mind that something like web3.storage or Estuary.tech may be better for your needs if you need easy access to the files later; Filecoin can (currently) be thought of as closer to archival storage, but efforts are being made to make it more generally useful too.