• 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
mrq closed issue mrq/tortoise-tts#4 2023-02-06 04:06:58 +00:00
do_tts.py and read.py
mrq pushed to main at mrq/tortoise-tts 2023-02-06 03:45:00 +00:00
a1f3b6a4da fixed up the computing conditional latents
mrq commented on issue mrq/tortoise-tts#4 2023-02-06 02:17:56 +00:00
do_tts.py and read.py

Use the Line Delimiter input in the web UI to process your text input into pieces, similar to read.py's behavior. For example, set Line Delimiter to \n for it to process each line one by…

mrq pushed to main at mrq/tortoise-tts 2023-02-06 00:56:28 +00:00
2cfd3bc213 updated README (before I go mad trying to nitpick and edit it while getting distracted from an iToddler sperging)
mrq pushed to main at mrq/tortoise-tts 2023-02-05 23:52:21 +00:00
945136330c Forgot to rename the cached latents to the new filename
mrq pushed to main at mrq/tortoise-tts 2023-02-05 23:25:48 +00:00
5bf21fdbe1 modified how conditional latents are computed (before, it just happened to only bother reading the first 102400/24000=4.26 seconds per audio input, now it will chunk it all to compute latents)
mrq pushed to main at mrq/tortoise-tts 2023-02-05 23:23:52 +00:00
6e89dcb97a modified how conditional latents are computed (before, it just happened to only bother reading the first 102400/24000=4.26 seconds per audio input, now it will chunk it all to compute latents)
mrq pushed to main at mrq/tortoise-tts 2023-02-05 23:20:54 +00:00
f19cbda183 modified how conditional latents are computed (before, it just happened to only bother reading the first 102400/24000=4.26 seconds per audio input, now it will chunk it all to compute latents)
mrq pushed to main at mrq/tortoise-tts 2023-02-05 20:10:56 +00:00
f66754b557 oops
mrq pushed to main at mrq/tortoise-tts 2023-02-05 18:00:10 +00:00
1c582b5dc8 added button to refresh voice list, enabling KV caching for a bonerific speed increase (credit to https://github.com/152334H/tortoise-tts-fast/)
mrq pushed to main at mrq/tortoise-tts 2023-02-05 17:42:04 +00:00
mrq pushed to main at mrq/tortoise-tts 2023-02-05 17:37:35 +00:00
c75e2a7431 enabling KV caching for a bonerific speed increase (credit to https://github.com/152334H/tortoise-tts-fast/)
mrq pushed to main at mrq/tortoise-tts 2023-02-05 14:46:15 +00:00
8831522de9 New tunable: pause size/breathing room (governs pause at the end of clips)
mrq pushed to main at mrq/tortoise-tts 2023-02-05 06:55:27 +00:00
c7f85dbba2 Fix to keep prompted emotion for every split line
mrq pushed to main at mrq/tortoise-tts 2023-02-05 06:42:00 +00:00
79e0b85602 Updated .gitignore (that does not apply to me because I have a bad habit of having a repo copy separate from a working copy)
mrq commented on issue mrq/tortoise-tts#3 2023-02-05 06:39:27 +00:00
Suggestion: Snap the "Candidates" slider to integer values

Added.

In bf32efe5038089beb0ccfbade50e290d886827fc :

  • had to fix a case with it breaking on candidates=1
  • did some cleanup for the printed messages
  • had the string to split by a…
mrq pushed to main at mrq/tortoise-tts 2023-02-05 06:35:49 +00:00
bc567d7263 Skip combining if not splitting, also avoids reading back the audio files to combine them by keeping them in memory
mrq pushed to main at mrq/tortoise-tts 2023-02-05 06:18:01 +00:00
bf32efe503 Added multi-line parsing
mrq commented on issue mrq/tortoise-tts#3 2023-02-05 05:53:59 +00:00
Suggestion: Snap the "Candidates" slider to integer values

However there's a problem in the latest commit where regardless of what sample and iteration values you set, it sets it to "512" during generation.

Already ahead of you, fixed it. When I was…

mrq pushed to main at mrq/tortoise-tts 2023-02-05 05:52:03 +00:00
cd94cc8459 Fixed accidentally not passing user-provided samples/iteration values (oops), fixed error thrown when trying to write unicode because python sucks