-
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
Progress report for whoever cares about my descent:
- scrapped my
-quartertests; it's not worth the time testing a gimped model if I want to validate that I can even get decent results with…
e59e4a9cfc
added another yaml parameter to only load the weights, effectively able to finetune (I'm multi-pass training this bitch)
[Training] [2023-03-26T08:45:43.861672] local_state[k] = v[grad_accum_step] [Training] [2023-03-26T08:45:43.861672] IndexError: list index out of range
47fdc78baa
added yaml params to set checkpoint save/load tag (set to it or step to save checkpoint name by it step number)
721dc9cfad
added accuracy metrics (requires torchmetrics), added evaluation loss (requires auraloss)
5d1414b7af
added accuracy metrics (requires torchmetrics), added evaluation loss (requires auraloss)
I suppose one last thing before I go back into my hole for another few days: training it is shaping up to be a real bitch and a half. I suppose it's only natural it is so, given it's a language…
Hello, I don't know if it's of any concern, but someone on the newer repository uploaded a trained model as detailed in this thread: https://github.com/lifeiteng/vall-e/issues/58
Neato.
…
ff5d47a952
something happened (something fucked up my AR, I don't know what broke it, but it's fixed after either of these)