From 8e5a79e11292c8a71b4718cca9b39ca073b7b689 Mon Sep 17 00:00:00 2001 From: collectivat Date: Tue, 7 Aug 2018 16:46:21 +0200 Subject: [PATCH] download link added --- README.md | 2 ++ scripts/setup_corpus.sh | 6 +++--- 2 files changed, 5 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index a055c1f..26dd816 100644 --- a/README.md +++ b/README.md @@ -64,6 +64,8 @@ la seva abraçada havia estat una batalla el clímax una victòria ### Corpus It _will be_ possible to download the audio corpus with the transcriptions and run `sphinxtrain`. For that, the script `setup_corpus.sh` is provided. It downloads a tar.gz archive which has the audio corpus with the necessary directory structure for `sphinxtrain`. Extracts it and writes the configuration file. +> **Note:** The `setup_corpus.sh` is not working yet. However the 240 hour acoustic corpus can be downloaded from [here](http://laklak.eu/share/tv3_0.3.tar.gz). + ``` $ source scripts/setup_corpus.sh --2018-03-01 16:51:23-- https://transfer.sh/ZPZ0C/ca-ca-0.1.gpg diff --git a/scripts/setup_corpus.sh b/scripts/setup_corpus.sh index 1fb8c45..1416f97 100644 --- a/scripts/setup_corpus.sh +++ b/scripts/setup_corpus.sh @@ -5,14 +5,14 @@ NOT_FOUND="ERROR: Could not find the sphinxtrain libraries locally. Make sure sp After installation rerun the script or replace the PROJECT_PATH and SPHINX_LIB_PATH variables manually in the PROJECT_PATH/etc/sphinx_train.cfg file." -MODEL="ca-ca-0.1" +MODEL="tv3_0.3" # download encrypted tar file -wget https://transfer.sh/ZPZ0C/$MODEL.gpg +wget http://laklak.eu/share/tv3_0.3.tar.gz #TODO check download # extract tar file -gpg --decrypt < $MODEL.gpg | tar xzf - +tar xzf tv3_0.3.tar.gz #TODO check directory # get local absolute path of the script