提交 3e5fc3dd 编写于 作者: J Junkun

Merge branch 'develop' of https://github.com/LittleChenCc/DeepSpeech into develop

......@@ -6,8 +6,7 @@
# This configuration is based on full-band MelGAN but the hop size and sampling
# rate is different from the paper (16kHz vs 24kHz). The number of iteraions
# is not shown in the paper so currently we train 1M iterations (not sure enough
# to converge). The optimizer setting is based on @dathudeptrai advice.
# https://github.com/kan-bayashi/ParallelWaveGAN/issues/143#issuecomment-632539906
# to converge).
###########################################################
# FEATURE EXTRACTION SETTING #
......@@ -136,4 +135,4 @@ eval_interval_steps: 1000 # Interval steps to evaluate the network
# OTHER SETTING #
###########################################################
num_snapshots: 10 # max number of snapshots to keep while training
seed: 42 # random seed for paddle, random, and np.random
\ No newline at end of file
seed: 42 # random seed for paddle, random, and np.random
......@@ -6,8 +6,7 @@
# This configuration is based on full-band MelGAN but the hop size and sampling
# rate is different from the paper (16kHz vs 24kHz). The number of iteraions
# is not shown in the paper so currently we train 1M iterations (not sure enough
# to converge). The optimizer setting is based on @dathudeptrai advice.
# https://github.com/kan-bayashi/ParallelWaveGAN/issues/143#issuecomment-632539906
# to converge).
###########################################################
# FEATURE EXTRACTION SETTING #
......
......@@ -57,7 +57,7 @@ else
echo "missing ${dec_conf}"
exit 1
fi
# NOTE(kan-bayashi): preprocess conf is optional
# preprocess conf is optional
if [ -n "${preprocess_conf}" ]; then
tar rfh ${outfile}.tar ${preprocess_conf}
echo -n " - preprocess config file: \`"
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册