shithub: opus

Info  •  Files  •  Log  •  Branches

brought osce/README.md up to date

--- a/dnn/torch/osce/README.md

+++ b/dnn/torch/osce/README.md

@@ -21,10 +21,13 @@

 which on top provides some random scaling. Data is taken from the datasets listed in dnn/datasets.txt and the exact list of items used for training and validation is

 located in dnn/torch/osce/resources.

-Second step is to run a patched version of opus_demo in the dataset folder, which will produce the coded output and add feature files. To build the patched opus_demo binary, check out the exp-neural-silk-enhancement branch and build opus_demo the usual way. Then run

+Second step is to run the opus_demo binary build with the --enable-osce-training-data flag, i.e. `./configure --disable-shared --enable-osce-training-data && make clean && make -j`

-`cd dataset && <path_to_patched_opus_demo>/opus_demo voip 16000 1 9000 -silk_random_switching 249 clean.s16 coded.s16 `

+which will enable writing decoder features to disk. To create the training data, then run

+

+`cd dataset && <path_to_opus_demo_with_training_data>/opus_demo voip 16000 1 9000 -silk_random_switching 249 clean.s16 coded.s16 `

+

 The argument to -silk_random_switching specifies the number of frames after which parameters are switched randomly.

 ## Regression loss based training

@@ -57,9 +60,16 @@

 to run it in background. In the latter case the output is written to `<output folder>/out.txt`.

+## Exporting a model to C

+To integrate a newly trained model into Opus, export model weights to opus/dnn by running

+

+`python export_model_weigths.py <path_to_checkpoint> ../../ [--quantize]`

+

+and then build opus with --enable-osce. Note, however that running autogen.sh again will overwrite any local changes to the model weights.

+

 ## Inference

 Generating inference data is analogous to generating training data. Given an item 'item1.wav' run

-`mkdir item1.se && sox item1.wav -r 16000 -e signed-integer -b 16 item1.raw && cd item1.se && <path_to_patched_opus_demo>/opus_demo voip 16000 1 <bitrate> ../item1.raw noisy.s16`

+`mkdir item1.se && sox item1.wav -r 16000 -e signed-integer -b 16 item1.raw && cd item1.se && <path_to_opus_demo_with_training_data>/opus_demo voip 16000 1 <bitrate> ../item1.raw noisy.s16`

 The folder item1.se then serves as input for the test_model.py script or for the --testdata argument of train_model.py resp. adv_train_model.py

-- 

⑨