Brandon Thomas
|
1e19067113
|
Repair model/ files that were overwritten for testing
|
2021-11-15 00:13:25 -05:00 |
|
Brandon Thomas
|
2b3fdb87e0
|
BROKEN: IP filter compiles, but does not block
|
2021-11-12 03:30:15 -05:00 |
|
Brandon Thomas
|
00131daedf
|
talknet downloads wip (BROKEN)
|
2021-09-10 05:16:58 -04:00 |
|
Brandon Thomas
|
da16826ab5
|
talknet works
|
2021-09-09 03:01:06 -04:00 |
|
Brandon Thomas
|
a681778046
|
end-to-end working? need to test the audio
|
2021-09-09 02:57:12 -04:00 |
|
Brandon Thomas
|
6db5f0187b
|
model loading, arpabet
|
2021-09-09 02:38:38 -04:00 |
|
Brandon Thomas
|
694789cef8
|
notes + requirements
|
2021-09-08 22:41:09 -04:00 |
|
Brandon Thomas
|
27c05a6207
|
create WIP talknet server
|
2021-09-08 22:35:58 -04:00 |
|
Brandon Thomas
|
c123d499ba
|
fix newlines issue
|
2021-09-03 02:41:42 -04:00 |
|
Brandon Thomas
|
2c8d512997
|
adjust relative imports
|
2021-08-29 05:05:47 -04:00 |
|
Brandon Thomas
|
584aedc905
|
clarifiying comment on requirements
|
2021-08-29 04:30:20 -04:00 |
|
Brandon Thomas
|
df36f9c246
|
closer to import working...
|
2021-08-29 04:28:53 -04:00 |
|
Brandon Thomas
|
6dfb8eca48
|
requirements for talknet (wip)
|
2021-08-29 04:24:35 -04:00 |
|
Brandon Thomas
|
340267b394
|
move talknet, etc. for reuse
|
2021-08-29 04:17:59 -04:00 |
|
Brandon Thomas
|
1a15029a22
|
reorganize tts models so pieces can be shared (BROKEN!)
|
2021-08-29 03:21:40 -04:00 |
|
Brandon Thomas
|
d70142219a
|
Move TalkNet out of tacotron2
|
2021-08-29 02:22:43 -04:00 |
|
Brandon Thomas
|
01a96f4141
|
Move NeMo/nemo out so it can be imported
|
2021-08-28 23:38:51 -04:00 |
|
Brandon Thomas
|
3839ae0dfc
|
Add __init__ files for module imports
|
2021-08-28 23:28:06 -04:00 |
|
Brandon Thomas
|
5c1d1a1bce
|
remove irrelevant requirements files
|
2021-08-28 23:18:47 -04:00 |
|
Brandon Thomas
|
fb877bc290
|
ControllableTalkNet from https://github.com/SortAnon/ControllableTalkNet
|
2021-08-28 23:15:29 -04:00 |
|
Brandon Thomas
|
f9871d9d05
|
Add NeMo from https://github.com/SortAnon/NeMo
|
2021-08-28 23:00:32 -04:00 |
|
Brandon Thomas
|
1fa3b9b880
|
When the w2l job fails, fail permanently
|
2021-08-28 02:35:00 -04:00 |
|
Brandon Thomas
|
c396a91881
|
Massive update + schema change. Add vocoder flexibility
|
2021-08-25 04:00:57 -04:00 |
|
Brandon Thomas
|
90e1db42c4
|
Sidecar supports both vocoders and uses the one in the HTTP request
|
2021-08-25 02:21:59 -04:00 |
|
Brandon Thomas
|
f4c8bb0095
|
plumbing for vocoders (mostly done)
|
2021-08-25 01:59:30 -04:00 |
|
Brandon Thomas
|
6ddc641475
|
intermediate plumbing for new vocoders (3)
|
2021-08-25 01:39:29 -04:00 |
|
Brandon Thomas
|
a61e144ecb
|
intermediate plumbing for new vocoders (2)
|
2021-08-25 01:32:37 -04:00 |
|
Brandon Thomas
|
21cffc1f03
|
intermediate plumbing for new vocoders
|
2021-08-25 01:11:36 -04:00 |
|
Brandon Thomas
|
c9112b7714
|
remove unused model code
|
2021-08-25 00:33:24 -04:00 |
|
Brandon Thomas
|
79ee6e408a
|
I think this works and sounds good!!
|
2021-08-22 21:28:18 -04:00 |
|
Brandon Thomas
|
d7c6e0d68d
|
almost working, but sounds super distorted
|
2021-08-22 21:26:11 -04:00 |
|
Brandon Thomas
|
3d61120ecb
|
this appears to work end-to-end, but I need to output the sound
|
2021-08-22 21:15:19 -04:00 |
|
Brandon Thomas
|
c52ce4de0a
|
WIP pipeline with hifigan
|
2021-08-22 21:08:01 -04:00 |
|
Brandon Thomas
|
a821afaebe
|
I can now load hifigan models from TT
|
2021-08-22 20:27:02 -04:00 |
|
Brandon Thomas
|
76fd658ae6
|
move hifigan to make import disambiguation easier
|
2021-08-22 20:07:43 -04:00 |
|
Brandon Thomas
|
e4fe7f8ed1
|
don't print devices
|
2021-08-22 19:42:07 -04:00 |
|
Brandon Thomas
|
6516a04eda
|
print devices
|
2021-08-22 19:41:50 -04:00 |
|
Brandon Thomas
|
d0ee1962d8
|
Move hifi-gan under tacotron2
|
2021-08-22 19:16:33 -04:00 |
|
Brandon Thomas
|
1a7fc2ab72
|
Add HifiGan model (https://github.com/SortAnon/hifi-gan.git)
|
2021-08-20 01:13:51 -04:00 |
|
Brandon Thomas
|
92949a719b
|
fix visibility db serialization bug
|
2021-08-07 18:51:09 -04:00 |
|
Brandon Thomas
|
b476485cd7
|
model loading/unloading works
|
2021-07-05 10:33:48 -04:00 |
|
Brandon Thomas
|
4da28003f2
|
attempt at running LRU for the sidecar
|
2021-07-05 02:18:20 -04:00 |
|
Brandon Thomas
|
5bc943a9f5
|
server seems to work!
|
2021-07-02 17:38:58 -04:00 |
|
Brandon Thomas
|
560beb0773
|
progress on loading models from multiple args
|
2021-07-02 17:27:02 -04:00 |
|
Brandon Thomas
|
321322c0a5
|
server is working and fast, but need to wire it up for API
|
2021-07-02 16:39:27 -04:00 |
|
Brandon Thomas
|
21ac34b8a5
|
wip
|
2021-07-02 04:21:31 -04:00 |
|
Brandon Thomas
|
3bbab7232c
|
Working on porting HTTP server from old project
7e33d9c7ca/cycle_gan/demo_server.py
|
2021-07-02 03:13:48 -04:00 |
|
Brandon Thomas
|
af80527c47
|
Check TT2 model for validity and store file size
|
2021-06-07 22:56:56 -04:00 |
|
Brandon Thomas
|
dfc52681cd
|
calculate audio metadata
|
2021-06-05 15:47:21 -04:00 |
|
Brandon Thomas
|
9c1a1e88f9
|
simplify mel generation
|
2021-06-05 15:36:30 -04:00 |