27024a7b38
Add oneAPI training support (ai-voice-cloning)
...
- Add an argument to use oneAPI when training
- Use it in the oneAPI startup
- Set an env var when doing so
- Initialize distributed training with ccl when doing so
Intel does not and will not support non-distributed training. I think that's a good decision.
The message that training will happen with oneAPI gets printed twice.
2023-05-04 20:51:31 +03:00
5092cf9174
Use ipexrun / better start
...
Supposedly, ipexrun improves performance by ~10%. It needs numactl.
Also deactivate any conda envs since working with oneAPI might mean having a conda env open.
This just reduces spam from setvars.sh.
2023-05-04 20:46:02 +03:00
c62d73f28a
Chmod start-oneapi.sh so it can actually be ran
2023-05-04 16:12:26 +03:00
7fedff0b1f
Make oneAPI startup its own file
...
Not sed-ing seems like a better idea, in case the user wants to revert it easily or something, or for updates.
2023-05-02 00:20:34 +03:00
7261ec2b1f
Use all custom wheels.
...
Intel's stable wheels that were released a day or so prior have a VRAM leak, or some other VRAM usage issue in general.
Also adds CCL, the stable wheels for which seem fine (?)
2023-05-01 23:19:10 +03:00
bd5e0c1d56
Add initial setup file
2023-04-30 19:13:25 +03:00
fd306d850d
updated setup-directml.bat to not hard require torch version because it's updated to torch2 now
2023-04-29 00:50:16 +00:00
eddb8aaa9a
indentation fix
2023-04-28 15:56:57 +00:00
99387920e1
backported caching of phonemizer backend from mrq/vall-e
2023-04-28 15:31:45 +00:00
c5e9b407fa
boolean oops
2023-04-27 14:40:22 +00:00
3978921e71
forgot to make the transcription tab visible with the bark backend (god the code is a mess now, I'll suck you off if you clean this up for me (not really))
2023-04-26 04:55:10 +00:00
b6440091fb
Very, very, VERY, barebones integration with Bark (documentation soon)
2023-04-26 04:48:09 +00:00
faa8da12d7
modified logic to determine valid voice folders, also allows subdirs within the folder (for example: ./voices/SH/james/ will be named SH/james)
2023-04-13 21:10:38 +00:00
02beb1dd8e
should fix #203
2023-04-13 03:14:06 +00:00
8f3e9447ba
disable diarize button
2023-04-12 20:03:54 +00:00
d8b996911c
a bunch of shit i had uncommited over the past while pertaining to VALL-E
2023-04-12 20:02:46 +00:00
mrq
b785192dfc
Merge pull request 'Make convenient to use with Docker' ( #191 ) from psr/ai-voice-cloning:docker into master
...
Reviewed-on: mrq/ai-voice-cloning#191
2023-04-08 14:04:45 +00:00
psr
9afafc69c1
docker: add training script
2023-04-07 23:15:13 +00:00
psr
c018bfca9c
docker: add ffmpeg for whisper and general cleanup
2023-04-07 23:14:05 +00:00
psr
d64cba667f
docker support
2023-04-07 21:52:18 +00:00
0440eac2bc
#185
2023-03-31 06:55:52 +00:00
9f64153a28
fixes #185
2023-03-31 06:03:56 +00:00
4744120be2
added VALL-E inference support (very rudimentary, gimped, but it will load a model trained on a config generated through the web UI)
2023-03-31 03:26:00 +00:00
9b01377667
only include auto in the list of models under setting, nothing else
2023-03-29 19:53:23 +00:00
f66281f10c
added mixing models (shamelessly inspired from voldy's web ui)
2023-03-29 19:29:13 +00:00
c89c648b4a
fixes #176
2023-03-26 11:05:50 +00:00
41d47c7c2a
for real this time show those new vall-e metrics
2023-03-26 04:31:50 +00:00
c4ca04cc92
added showing reported training accuracy and eval/validation metrics to graph
2023-03-26 04:08:45 +00:00
8c647c889d
now there should be feature parity between trainers
2023-03-25 04:12:03 +00:00
fd9b2e082c
x_lim and y_lim for graph
2023-03-25 02:34:14 +00:00
9856db5900
actually make parsing VALL-E metrics work
2023-03-23 15:42:51 +00:00
69d84bb9e0
I forget
2023-03-23 04:53:31 +00:00
444bcdaf62
my sanitizer actually did work, it was just batch sizes leading to problems when transcribing
2023-03-23 04:41:56 +00:00
a6daf289bc
when the sanitizer thingy works in testing but it doesn't outside of testing, and you have to retranscribe for the fourth time today
2023-03-23 02:37:44 +00:00
86589fff91
why does this keep happening to me
2023-03-23 01:55:16 +00:00
0ea93a7f40
more cleanup, use 24KHz for preparing for VALL-E (encodec will resample to 24Khz anyways, makes audio a little nicer), some other things
2023-03-23 01:52:26 +00:00
d2a9ab9e41
remove redundant phonemize for vall-e (oops), quantize all files and then phonemize all files for cope optimization, load alignment model once instead of for every transcription (speedup with whisperx)
2023-03-23 00:22:25 +00:00
19c0854e6a
do not write current whisper.json if there's no changes
2023-03-22 22:24:07 +00:00
932eaccdf5
added whisper transcription 'sanitizing' (collapse very short transcriptions to the previous segment) (I really have to stop having several copies spanning several machines for AIVC, I keep reverting shit)
2023-03-22 22:10:01 +00:00
736cdc8926
disable diarization for whisperx as it's just a useless performance hit (I don't have anything that's multispeaker within the same audio file at the moment)
2023-03-22 20:38:58 +00:00
aa5bdafb06
ugh
2023-03-22 20:26:28 +00:00
13605f980c
now whisperx should output json that aligns with what's expected
2023-03-22 20:01:30 +00:00
8877960062
fixes for whisperx batching
2023-03-22 19:53:42 +00:00
4056a27bcb
begrudgingly added back whisperx integration (VAD/Diarization testing, I really, really need accurate timestamps before dumping mondo amounts of time on training a dataset)
2023-03-22 19:24:53 +00:00
b8c3c4cfe2
Fixed #167
2023-03-22 18:21:37 +00:00
da96161aaa
oops
2023-03-22 18:07:46 +00:00
f822c87344
cleanups, realigning vall-e training
2023-03-22 17:47:23 +00:00
909325bb5a
ugh
2023-03-21 22:18:57 +00:00
5a5fd9ca87
Added option to unsqueeze sample batches after sampling
2023-03-21 21:34:26 +00:00
9657c1d4ce
oops
2023-03-21 20:31:01 +00:00
