7639ca4b19
Update MUSICGEN.md
2023-10-27 15:06:44 +02:00
f73b7ae4f3
Many bug fixes for v1.1.0 ( #335 )
...
* wip
* fix bug in sample storage
* fix bug in fsdp with pytorch 2.0.1
* fix kld
* revert back passt change
* fix n_quantizers
* changes
* adding warning
* extra warnings and tests
* missing changelog
* missing changes between audiogen and musicgen
2023-10-26 17:58:04 +02:00
5d8752d356
changing tmpdir when running in slurm + not depending anymore on torchaudio for writing audio files. ( #306 )
...
* changing tmpdir when runnign in slurm
* fixing typing in dadam
* limiting dependency on torchaudio for writing files
* not using torchaudio for reading anymore
* trying desperatly to get those unit tests to pass
* plop
* fixing tests once more
* linter
* plop
2023-10-12 10:39:58 +02:00
a2b9675695
preparing v1.0.0 release ( #269 )
2023-09-07 18:03:13 +02:00
80def9cd3c
update SISNR doc to make it clear we display minus the SISNR ( #268 )
...
* doc
* making doc clearer
* line too long
2023-09-07 15:50:06 +02:00
3dea2b9832
Merge pull request #257 from facebookresearch/readme/cache_dir
...
Update README
2023-08-29 12:07:57 +02:00
b50c9a490e
Merge pull request #245 from starburst997/readme-ffmpeg
...
Use double quotes for better compatibility (REAMDE / ffmpeg)
2023-08-29 10:11:47 +02:00
9c405e0ecc
Smaller t5 model for debugging
2023-08-28 07:07:28 -07:00
b2a3660ce7
Update README
2023-08-28 06:40:34 -07:00
4eb194f394
Add transformers hub usage mbd ( #202 )
...
* add transformers hub usage for mbd models
* small bug nit correction
* correct style
2023-08-28 15:35:06 +02:00
656a49e4e7
Use double quotes for better compatibility
2023-08-28 09:30:17 -04:00
7cee8522e5
Add missing arguments to docstring ( #254 )
2023-08-28 14:19:38 +02:00
5ecdc68ef7
Merge pull request #251 from 0xlws/clone
...
fix: add missing protobuf dependency
2023-08-25 13:12:20 +02:00
8554c30b12
fix: add missing protobuf to requirements.txt
2023-08-25 11:45:20 +02:00
e6023d3ded
cleanup: rm unneccesary torch import
2023-08-25 11:44:32 +02:00
e96018613a
Merge pull request #182 from carankt/main
...
Right call for Encodec Model
2023-08-12 09:13:03 +02:00
807a96cafc
Merge pull request #214 from facebookresearch/fixes_mbd
...
Fixes mbd
2023-08-12 09:08:23 +02:00
ac311afe52
add device
2023-08-11 08:57:38 -07:00
b9489711d8
add sample page
2023-08-11 08:55:20 -07:00
1b4012ab24
Merge pull request #213 from facebookresearch/update_kld
...
Update KLD with PaSST to reproduce MusicGen results
2023-08-11 10:50:18 +03:00
b960e693fc
Update comment in Makefile
2023-08-10 14:37:18 +00:00
7a5a4b38d5
Update model card
2023-08-10 14:26:32 +00:00
2f4f9248b2
Update AudioGen KLD PaSST to 10sec in eval config
2023-08-10 13:55:01 +00:00
f26fa4466f
Update model card
2023-08-09 15:24:55 +00:00
2454f809f4
Update KLD with PaSST to reproduce MusicGen results
2023-08-09 13:42:49 +00:00
d6df4f0fff
Merge pull request #168 from srezasm/main-1
...
Fix: Use facebook/musicgen-small instead of facebook/musicgen-large
2023-08-08 11:00:22 +02:00
f1ee52fa15
MBD kept loading on each generation ( #184 )
2023-08-07 16:11:12 +02:00
1048255da9
add warning
2023-08-07 11:26:59 +02:00
3df5daff72
automatically putting dac into eval model
2023-08-07 10:25:22 +02:00
b8334656f8
Merge pull request #1 from carankt/carankt-patch-1
...
Update Command to work with EncoDec
2023-08-04 10:25:38 -06:00
163b61f8a7
Update Command to work with EncoDec
...
Loading the encodec model requires the call to be "encodec_model_24khz()" instead of the "get_*_()".
2023-08-04 10:24:36 -06:00
c09cd0cb90
Merge pull request #164 from frinkleko/main
...
typo(musicgen_app.py):'ou can' -> 'you can'
2023-08-04 11:33:33 +02:00
c96eef5eb7
Merge pull request #159 from bocytko/patch-1
...
AudioGen README: Removes call to non-existing method 'generate_unconditional'
2023-08-04 11:12:35 +02:00
3e7eb128e8
Fix: Use facebook/musicgen-small instead of facebook/musicgen-large
2023-08-03 15:30:02 +03:30
dd30dfeed4
typo(musicgen_app.py):'ou can' -> 'you can'
2023-08-03 14:02:57 +08:00
efacc4879b
Removes call to non-existing generate_unconditional
...
When running the example, one gets an error:
```
'AudioGen' object has no attribute 'generate_unconditional'
```
AudioGen does not provide a generate_unconditional method. The example looks like a copy-paste mistake from Musicgen readme.
2023-08-03 00:19:27 +02:00
c5157b5bf1
Update README for MBD
2023-08-02 16:39:21 +02:00
24a5c75e42
Update README
2023-08-02 15:41:56 +02:00
ec3df02e89
Update team config
2023-08-02 15:40:33 +02:00
5e3ae32f69
Update LICENSE_weights
...
nicer formatting for CC NC BY
2023-08-02 15:34:38 +02:00
bf7052227b
AudioCraft v1.0.0 release with training code, AudioGen, MultiBandDiffusion etc.
2023-08-02 15:14:53 +02:00
539ec044a2
fix updated linter
2023-08-01 20:33:11 +02:00
207888c76a
v0.0.2
2023-08-01 20:20:15 +02:00
6b23f77215
update colab
2023-07-14 12:16:49 +02:00
d874966c6c
[README] Add section on 🤗 Transformers ( #135 )
...
* [README] Add section on 🤗 Transformers
* fix docs link
2023-06-30 12:32:12 +02:00
f187760a31
linting and adding file cleaner
2023-06-29 16:28:22 +02:00
7e2f6d601c
Update README.md
2023-06-22 16:53:29 +02:00
7ccdf69e6e
Model can be loaded from local directory ( #69 )
...
* Model can be loaded from local directory
* applied flake8 linter
2023-06-22 16:49:18 +02:00
6b3bb5374c
add option to record from mic ( #87 )
2023-06-22 16:35:07 +02:00
c9179f87d9
again
2023-06-13 18:46:00 +03:00
Alexandre Défossez