Help with Index Training

[hearty basin]

Im literally run collab how i did before, no changes and now i got this error

other cells runs perfect without errors

[formal imp]

You screwed up on dataset formatting

The dataset must be formatted like this:

• folder with your model's name (must be the same as experiment_name you will set on the Colab, no spaces too)
• Audios here (no spaces here too)

So in this case, the name of the folder you zipped should be "LEGOClone".

Also, make sure that you've exported the audios from Audacity (or RX10?) and that you zipped it as .zip, and not .rar

[hearty basin]

Yeah it is

datasets are from game untouched

[formal imp]

I think you should import them into Audacity and export them as one wav file

[hearty basin]

dammn

[formal imp]

just to prevent errors yk

[hearty basin]

im already trained model for 800 epoch is there any quick way to fix index?

[hearty basin]

or i dont need to train 800 epoches again and just fix files and train only index?

[formal imp]

yeah uh, no

you'll have to train again cuz preprocessing and feature extraction prolly screwed up

and therefore, the pths generated are also terrible, probably

[hearty basin]

aaa damn...

ok thank you!!

[formal imp]

you're welcome, lmk if the issue is fixed

[hearty basin]

ok

[paper maple]

uhh bro

it says if he tried feature extraction

he didnt run the feature extraction thing

says on the bottom

[formal imp]

but, that's the problem

most people do

BUT

it doesn't do its thing

(it says no-feature-todo, if you're wondering)

and that just screws up the model

[paper maple]

ohhh alright i mean ive never had an issue bc im a payer 🤓 i pay as i go using the colab and use a different one

i mean i got 6 different alts

and i use the disconnected

but alr alr

[formal imp]

...nice i guess

[paper maple]

https://tenor.com/view/my-reaction-reaction-my-reactoin-to-my-reaction-to-that-my-reaction-to-that-information-gif-27615079