-
Notifications
You must be signed in to change notification settings - Fork 150
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[DOCUMENTATION] Tutorial for training using arbitrary reamping data #371
Comments
Is there a guide for what type of recording to use to train the model? Do you just record your guitar through the amp or what? |
@DimGakis that sounds pretty hacky. Was the file you used based on v1_1_1.wav? Or how is it related to one of the standard reamp files? (Assuming it is). If it's not related, then you need to use the CLI trainer. |
But the fact that it doesn't "soft match" any of the standardized files means that something is different in it that makes it not have something that it needed to have. |
My own test tone is five minuets long and is not related to any of the standard reamp file. With the old trainer sightly modified I was able to use my own test tone and create some great sounding profiles. I don't know what is the CLI trainer, how to find it or how to use it. If its hard to help me with the CLI trainer then can you please tell me how I can get older versions of the trainer so I can create the profiles that are recorded with my own test tone the way I used to? I can not redo the recordings with the supported reamp files because I don't own the equipment anymore. |
Directions for using the CLI trainer: https://github.com/sdatkinson/neural-amp-modeler/tree/main#the-command-line-trainer-all-features |
Updated now that there's a ReadTheDocs (hooray!): https://neural-amp-modeler.readthedocs.io/en/latest/tutorials/command-line.html I'm thinking to add a tutorial for how to use NAM if you have a DI and the reamp, but you can't reamp one of the standardized audio files (e.g. you want to make a NAM of something you recorded in the past). I'm going to change this Issue to track making that tutorial page for the documentation. |
On a previous version of the trainer I used a custom test tone that I created by myself and produced great profiles.
The only thing I needed to do was to modify the 'core.py' file that's found in 'neural-amp-modeler\nam\train'
so the :
'input_version: Optional[Version] = None,'
Was replaced with:
'input_version: Optional[Version] = Version(1, 1, 1),'
Now that I updated to the current version of the trainer this modification is not working and I have many recordings that are done using my mentioned test tone.
Some of the amps were rented so It's impossible to redo the recordings with the supported test tones.
Can someone please guide me on how I can modify the current trainer to accept my own test tone like the old one did?
If something like that is impossible in the current version, can someone tell me where to find the old trainer ?
Please keep in mind that I am not an expert and I will need some detailed guidance.
Kind regards.
The text was updated successfully, but these errors were encountered: