David
|
ea160a8b39
|
debugging Travis run of ctests
|
2019-09-29 09:56:16 +09:30 |
David
|
0c0ba329fc
|
debugging Travis run of ctests
|
2019-09-29 09:16:42 +09:30 |
David
|
3ac5cee235
|
ctests pasing again
|
2019-09-29 08:50:57 +09:30 |
David
|
fb2d791b2e
|
use 190924a tar ball as default
|
2019-09-29 07:52:02 +09:30 |
David
|
69bdcfbd06
|
cleaned up dump data a little
|
2019-09-29 07:00:00 +09:30 |
David
|
02ecf2ac30
|
190924a, using original LPCNet voicing est (pitch gain). I have accidentally written over the 190922a network with c2voicing, so will run again
|
2019-09-24 19:00:44 +09:30 |
David
|
49c95df13d
|
cq_16kHz fine, peter a bit spikey, slightly better than 190920c, but not as gd at 190920a. However, quite remarkably - all.wav sounds fine, even outliers like canadian. This is puzzling, as this training material doesn't include resampling, or wianews sample
|
2019-09-23 06:51:26 +09:30 |
David
|
24bc72efef
|
190920c in between 190215 and 190920a, not quite good enough
|
2019-09-22 14:45:44 +09:30 |
David
|
18c4f2ba33
|
just peter and david, but cepstrals rather than mags, sounds as good as previous with mags
|
2019-09-20 11:27:54 +09:30 |
David
|
a2df6288a9
|
training script with just Peter and David that produces improvements over the 190215 network
|
2019-09-20 07:34:34 +09:30 |
David
|
f86b8da270
|
about to reconfigure for just vk5apr and david
|
2019-09-19 05:36:29 +09:30 |
drowe67
|
8b78d3f5bc
|
Merge pull request #12 from ra1nb0w/cmake
cmake: add DISABLE_CPU_OPTIMIZATION option
Based on Richard's review I'll merge. Thanks.
|
2019-09-02 15:29:41 +09:30 |
Davide Gerhard
|
604021c814
|
cmake: add DISABLE_CPU_OPTIMIZATION option
add new option to disable cpu flags auto-discovery.
needed by binary package manager because not all architecture has avx.
|
2019-08-21 09:29:04 +02:00 |
drowe67
|
0d5292c4dd
|
Merge pull request #11 from ra1nb0w/cmake
fix cmake configuration file name
|
2019-08-19 05:43:11 +09:30 |
Davide Gerhard
|
b9ac3c74dc
|
fix cmake configuration file name
fix installed cmake configuration as requested by freedv-gui
|
2019-08-12 18:05:49 +02:00 |
David
|
0693c3c3db
|
added option to use codec 2 voicing estimator. This works well on tinytrain.sh with single m and f speaker
|
2019-08-04 19:03:31 +09:30 |
David
|
a67030ba51
|
using 190804a as the mag unittest, to presever network
|
2019-08-04 15:47:31 +09:30 |
David
|
6ac9b1902d
|
finally - high quality from tinytrain on m & f samples, samme quality and loss as 4a3998
|
2019-08-04 15:18:58 +09:30 |
David
|
45d70671d1
|
added log mag support and test
|
2019-08-03 08:19:02 +09:30 |
David
|
547b6f65f9
|
added log mag support and test
|
2019-08-03 08:18:54 +09:30 |
David
|
286535a391
|
adding NN files in f32 format
|
2019-08-03 06:45:28 +09:30 |
David
|
4c20759367
|
debugging travis
|
2019-08-02 16:35:11 +09:30 |
David
|
7cc35231ab
|
first pass a ctests and Travis support
|
2019-08-02 16:16:38 +09:30 |
David
|
85df2d36e7
|
synth and synth_20h tests
|
2019-08-02 15:36:35 +09:30 |
David
|
d0998f478e
|
refactoring
|
2019-08-02 15:26:39 +09:30 |
David
|
bf46e5fa2b
|
run time load of NNs working, building up tests for two nets
|
2019-08-02 15:12:53 +09:30 |
David
|
ab81ee805a
|
dump_lpcnet generates constant size sparse arrays
|
2019-08-02 12:23:58 +09:30 |
David
|
cff1d4a888
|
tested and fixed bug in error detection
|
2019-08-02 11:32:37 +09:30 |
David
|
ea89bd51d8
|
included octvae comparison of states
|
2019-08-02 11:12:22 +09:30 |
David
|
63d1f4d148
|
basic synth test passing
|
2019-08-02 09:18:43 +09:30 |
David
|
d24c4bdf63
|
added nnet rw files
|
2019-08-02 08:40:11 +09:30 |
drowe67
|
d97a352913
|
Update README.md
|
2019-06-16 05:04:09 +09:30 |
drowe67
|
a7eb95eb5e
|
update after codec2/brad-2020 merged
|
2019-05-29 07:13:58 +09:30 |
drowe67
|
50049fd599
|
Update README.md
|
2019-05-28 20:50:04 +09:30 |
David
|
420699ca14
|
removed occasional printf
|
2019-05-17 17:30:51 +09:30 |
drowe67
|
c946f616b0
|
Merge pull request #7 from drowe67/dr-file-io-for-windows
Debugging LPCNet on Windows
|
2019-05-17 16:20:05 +09:30 |
David Rowe
|
83eb6623a4
|
make sure we don't supply out of range pitch values with bit errors
|
2019-05-17 13:07:54 +09:30 |
David Rowe
|
12e83060b3
|
clean up and README notes on cross compiling for Windows
|
2019-05-14 18:43:14 +09:30 |
David Rowe
|
43bf0950f9
|
traced click problem to PRNG rteunfign RAND_MAX, fixed now, and lpcnet_enc/lpcnet_dec work OK on Windows (wine) without clicks
|
2019-05-14 06:44:16 +09:30 |
David Rowe
|
7878ce9074
|
have both versions outputting the same values, but now both clicking! Key change (I think) was a 32 bit rand() function...
|
2019-05-13 21:20:36 +09:30 |
David Rowe
|
4eb5ca57fa
|
sorting out pitch confusion
|
2019-05-13 06:17:49 +09:30 |
David Rowe
|
43a6352353
|
progressing Wondows click problem, found tdump and dump_data were outputing pitch indexes of 320 which is out of range of pitch embed layer input of 0..255. frame rate vectors from Linux and Windows versions now match, and clicks gone from tdump | test_lpcnet test.
|
2019-05-12 11:23:06 +09:30 |
David Rowe
|
ed9494dead
|
building up test framework to track down clicking bug on Windows
|
2019-05-12 07:29:26 +09:30 |
David Rowe
|
645ab41410
|
windows likes specific file type flags with fopen
|
2019-05-11 06:14:44 +09:30 |
David Rowe
|
2af267387f
|
throw away test program
|
2019-05-10 15:00:25 +09:30 |
David
|
2bad2de563
|
added --infile --outfile as stdin/stdout pipes are tricky under Windows/Wine
|
2019-05-08 06:39:49 +09:30 |
Richard Shaw
|
53c4bec885
|
Need to improve CPU optimization logic but it works for now.
|
2019-04-23 14:48:06 -05:00 |
Richard Shaw
|
823ec939e2
|
Merge branch 'master' of github.com:drowe67/LPCNet
|
2019-04-23 13:23:03 -05:00 |
Richard Shaw
|
ae9868de64
|
Add install targets for lpcnet_{enc,dec}.
|
2019-04-23 13:22:55 -05:00 |
Richard
|
e54814ad69
|
Merge pull request #5 from darksidelemm/osx_fixes
Small changes to enable compilation under OSX
|
2019-04-22 20:18:54 -05:00 |