More Clarity on the Dataset Uploading

[sohamsshah]

I am super excited to contribute to this Project. I went through the project details, concept and found it a helpful tool for AI in future. Just wanted to clarify my understanding about the issue #113 which I am working on. And seeing this project, I see many potential similar work to be done.

Steps I followed:
I have to set up the dev env.
Download the corresponding dataset from the origin website that is to be uploaded on my local.
Write code to read that dataset in numpy arrays.
Use .load method to load it to the cloud.
Push the code(not the dataset which is on my local machine) to github and create a PR.

Have I understood correctly?

If this is the case then,
Would be super glad to contribute and also suggest changes in documentation for more clear idea of this task.

[AbhinavTuli]

Hey @sohamsshah. It's good to see you excited to work on our project!
You've almost nailed the steps. A couple of minor corrections:-

• You need to store the dataset onto the hub platform using ds.store() and not load
• You then need to verify the integrity of the uploaded dataset by loading it using hub.load()
• Also verify that the dataset is visualizable on app.activeloop.ai

Feel free to ask for more clarifications if you still have doubts

[aksh-02]

Thanks. This makes things little more easier.