Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

small updates

This commit is contained in:
erogol 2020-10-26 17:16:16 +01:00
parent 741d278cac
commit e982156009
2 changed files with 15 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
TTS/bin/train_wavernn_vocoder.py
View File

@ -95,7 +95,6 @@ def train(model, optimizer, criterion, scheduler, ap, global_step, epoch):
end_time = time.time() end_time = time.time()
c_logger.print_train_start() c_logger.print_train_start()
# train loop # train loop
print(" > Training", flush=True)
for num_iter, data in enumerate(data_loader): for num_iter, data in enumerate(data_loader):
start_time = time.time() start_time = time.time()
x_input, mels, y_coarse = format_data(data) x_input, mels, y_coarse = format_data(data)

6
TTS/vocoder/configs/wavernn_config.json
View File

@ -41,7 +41,7 @@
// }, // },
// MODEL MODE // MODEL MODE
"mode": 10, // mold [string], gauss [string], bits [int] "mode": "mold", // mold [string], gauss [string], bits [int]
"mulaw": true, // apply mulaw if mode is bits "mulaw": true, // apply mulaw if mode is bits
// MODEL PARAMETERS // MODEL PARAMETERS
@ -58,7 +58,7 @@
// DATASET // DATASET
//"use_gta": true, // use computed gta features from the tts model //"use_gta": true, // use computed gta features from the tts model
"data_path": "/media/alexander/LinuxFS/SpeechData/GothicSpeech/NPC_Speech", // path containing training wav files "data_path": "/home/erogol/Data/LJSpeech-1.1/wavs/", // path containing training wav files
"feature_path": null, // path containing computed features from wav files if null compute them "feature_path": null, // path containing computed features from wav files if null compute them
"seq_len": 1280, // has to be devideable by hop_length "seq_len": 1280, // has to be devideable by hop_length
"padding": 2, // pad the input for resnet to see wider input length "padding": 2, // pad the input for resnet to see wider input length
@ -93,5 +93,5 @@
"eval_split_size": 50, // number of samples for testing "eval_split_size": 50, // number of samples for testing
// PATHS // PATHS
"output_path": "output/training/path" "output_path": "/home/erogol/Models/LJSpeech/"
} }