mrq
18 Followers · 0 Following

  • 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
Repositories
15
 Projects Packages Public Activity Starred Repositories
mrq commented on issue mrq/ai-voice-cloning#343 2023-08-24 15:24:59 +00:00
Vall-E Backend Training: RuntimeError: Failed to find any .qnt.pt file in specified training dataset.

Fixed in commit 22904a86

mrq pushed to master at mrq/vall-e 2023-08-24 15:24:20 +00:00
22904a8639 more oversights fixed because I've been using a cached dataloader forever now and didn't catch these problems
mrq commented on issue mrq/ai-voice-cloning#343 2023-08-24 14:19:54 +00:00
Vall-E Backend Training: RuntimeError: Failed to find any .qnt.pt file in specified training dataset.

What a silly oversight. Fixed in commit 5873c27f1ad2094b42489401ad961bbc8c8b7ffb.

mrq pushed to master at mrq/vall-e 2023-08-24 14:19:37 +00:00
5873c27f1a ops
mrq commented on pull request mrq/ai-voice-cloning#341 2023-08-24 12:30:25 +00:00
fix filename generation which didn't work and overwrote existing files

Right, I kept forgetting to try and fix it in my VALL-E inference tests.

mrq pushed to master at mrq/ai-voice-cloning 2023-08-24 12:30:04 +00:00
4aa240d48a Merge pull request 'fix filename generation which didn't work and overwrote existing files' (#341) from ben_mkiv/ai-voice-cloning:master into master
00b173857d fix filename generation which didn't work and overwrote existing files
Compare 2 commits »
mrq merged pull request mrq/ai-voice-cloning#341 2023-08-24 12:30:02 +00:00
fix filename generation which didn't work and overwrote existing files
mrq pushed to master at mrq/ai-voice-cloning 2023-08-23 22:09:30 +00:00
dc46fdc7d0 fixed another issue from haphazardly copying my changes from my training machine
mrq pushed to master at mrq/vall-e 2023-08-23 22:02:15 +00:00
501a857d5d ops
mrq commented on issue mrq/ai-voice-cloning#339 2023-08-23 21:52:39 +00:00
Vall-E Backend Training: "list indices must be integers or slices, not dict"

The root issue should be fixed in commit 29290f574eb2f8aa9c2cf19b9aaf131a2fd1c3ff.

Additionally, if you were going to finetune with the web UI, generating the training YAML should be working…

mrq pushed to master at mrq/ai-voice-cloning 2023-08-23 21:49:27 +00:00
29290f574e should fix issue that arises when trying to prepare the dataset without slicing segments
mrq pushed to master at mrq/ai-voice-cloning 2023-08-23 21:48:29 +00:00
25625d5049 should fix issue that arises when trying to prepare the dataset without slicing segments
mrq commented on issue mrq/ai-voice-cloning#335 2023-08-23 21:46:02 +00:00
(Prepare Dataset) List indices must be intergers or slices, not dict

Odd though, as I never needed to before haha.

I'm very, very certain it was a regression I've introduced when committing the changes from my training rig back to this repo (commit 72a38ff2fc6a…

mrq pushed to master at mrq/ai-voice-cloning 2023-08-23 21:42:47 +00:00
0a5483e57a updated valle yaml template
mrq pushed to master at mrq/vall-e 2023-08-23 21:42:00 +00:00
4585824cd3 tweaks, including exporting on save/quit
mrq commented on pull request mrq/ai-voice-cloning#336 2023-08-23 21:06:48 +00:00
favor existing arguments from parameters (kwargs) over global (args)

ignore the commit spam... squashed it down to one commit

No worries.

mrq merged pull request mrq/ai-voice-cloning#336 2023-08-23 21:05:41 +00:00
favor existing arguments from parameters (kwargs) over global (args)
mrq pushed to master at mrq/ai-voice-cloning 2023-08-23 21:05:41 +00:00
e613299304 Merge pull request 'favor existing arguments from parameters (kwargs) over global (args)' (#336) from ben_mkiv/ai-voice-cloning:master into master
ce24ba41e2 Websocket server, override args parameters for model settings (squashed)
Compare 2 commits »
mrq commented on issue mrq/ai-voice-cloning#335 2023-08-23 21:03:53 +00:00
(Prepare Dataset) List indices must be intergers or slices, not dict

I haven't gotten a chance to actually look at it, but from the other issue I referenced to, you might need to chick Slice Segments, and set your offsets to -0.05 and 0.2 for safety. I…

mrq commented on issue mrq/ai-voice-cloning#339 2023-08-23 21:01:57 +00:00
Vall-E Backend Training: "list indices must be integers or slices, not dict"

I guess that would explain mrq/ai-voice-cloning#335 better. I'll finagle with the web UI and see why it's breaking when not slicing.

In the meantime you should…