-
https://git.ecker.tech/ aims to provide a place to share my efforts while maintaining true ownership of my code, as I do not trust GitHub.
XMR: 4B9TQdkAkBFYrbj5ztvTx89e5LpucPeTSPzemCihdDi9EBnx7btn8RDNZTBz2zihWsjMnDkzn5As1LU6gLv3KQy8BLsZ8SG
- Joined on
2022-10-10
It took some wrestling, but this seems the best I'll get: https://vocaroo.com/1eoBNEHDf4Uy
It's unironically just the default settings with half-precision enabled for generating. I don't…
I assume the webUI will automatically look to see if there is a set of finetune files for it, and use them if they exist in the training folder?
You'll need to go into Settings and select…
I'll poke around and check the outputs. desu, finetuned models seems to have an inherent slight loss in the output quality, unless it's trained very, very slowly and properly.
The things I can…
After playing around with it for a bit, the absolute simplest implementation would be adding api_name='generate' to the .click(.....) call here:
That's pretty much what I did earlier just to…
If I then start the training, it will (eventually) output ... some other files
Of (some) importance to the average user, it'll spit out:
- in
./trainings/{name}-finetune/models/, the final…
I kept meaning to share mine, but I kept deleting my finetunes thinking I can train them better after refining the training process.
I'll share mine on a [huggingspace repo](https://huggingface.…
Got it to replicate, the epoch counter manages to desync when it takes one step to complete an epoch.
I'll try and probe why that's so, despite having a line to re-sync.
Fixed in commit 206a14fdbe978afd46aa3f0f8414b6a7c5466cbf.
Oops, typo. Fixed in commit b82961ba8a9daf475562c86a96c769980ff489d5.