From 45a50d2dbaa35ed79f53afa10b0023ea3cd89a09 Mon Sep 17 00:00:00 2001
From: Michael Chung <michael.chung@databites.ca>
Date: Mon, 23 Oct 2023 19:58:25 -0400
Subject: [PATCH] Made some fixes to the container.

---
 tts/VALL-E-X/main.py                | 39 ++++++++++++++++++++---------
 tts/VALL-E-X/utils/prompt_making.py |  1 +
 2 files changed, 28 insertions(+), 12 deletions(-)

diff --git a/tts/VALL-E-X/main.py b/tts/VALL-E-X/main.py
index e42fec5..67998e2 100755
--- a/tts/VALL-E-X/main.py
+++ b/tts/VALL-E-X/main.py
@@ -150,22 +150,22 @@ voice_designer = VoiceDesigner()
 
 def main(args):
     
-    print(args.mode)
+    print("Starting Inference on Vall-E-X With Inputs")
+    print(f"Mode: {args.mode}")
     
-    print(args.text)
-    print(args.audio_wav_files)
-    print(args.whisper_folder_path)
-    print(args.vocos_folder_path)
+    print(f"Text: {args.text}")
+    print(f"List of Files For Create: {args.audio_wav_files}")
+    print(f"Whisper Folder Path: {args.whisper_folder_path}")
+    print(f"Vocos Folder Path: {args.vocos_folder_path}")
+    print(f"Vall-E-X Path: {args.vallex_path}")
     
-    print(args.vallex_path)
+    print(f"Prompt Path: {args.prompt_path}")
+    print(f"Prompt Name: {args.prompt_name}")
     
-    print(args.prompt_path)
-    print(args.prompt_name)
+    print(f"Audio Name: {args.audio_name}")
+    print(f"Audio Path: {args.audio_path}")
     
-    print(args.audio_name)
-    print(args.audio_path)
-    
-    print(args.tmp_work_dir)
+    print(f"Temp Work Dir: {args.tmp_work_dir}")
     
     voice_designer.temp_path = pathlib.Path(args.tmp_work_dir)
 
@@ -174,6 +174,7 @@ def main(args):
                                              pathlib.Path(args.whisper_folder_path))
     
     if args.mode == 0: # run inference
+        print("Running Inference")
         voice_designer.tts_with_prompt(prompt_dir=pathlib.Path(args.prompt_path),
                                        prompt_name=pathlib.Path(args.prompt_name),
                                        audio_output_path=pathlib.Path(args.audio_path),
@@ -234,3 +235,17 @@ if __name__ == "__main__":
 
 # create voice example
 #/home/tensor/code/storyteller/storyteller-ml/tts/VALL-E-X/.venv/bin/python /home/tensor/code/storyteller/storyteller-ml/tts/VALL-E-X/main.py --mode 1 --audio-wav-files "/home/tensor/code/TTSDockerContainer/Vall-E-mount/input/20.wav" "/home/tensor/code/TTSDockerContainer/Vall-E-mount/input/21.wav" --whisper-folder-path "/home/tensor/code/storyteller/storyteller-ml/tts/VALL-E-X/whisper" --whisper-model medium --vocos-folder-path "/home/tensor/code/TTSDockerContainer/Vall-E-mount/models/vocos-encodec-24khz" --vallex-path "/home/tensor/code/TTSDockerContainer/Vall-E-mount/models/vallex-checkpoint.pt" --prompt-path "/home/tensor/code/storyteller/VALL-E-X-TTS-Container/Vall-E-mount/prompts" --prompt-name "test_prompt" --tmp-work-dir "/tmp"
+
+#/home/tensor/code/storyteller/storyteller-ml/tts/VALL-E-X/.venv/bin/python 
+#/home/tensor/code/storyteller/storyteller-ml/tts/VALL-E-X/main.py 
+# --mode 1 
+# --audio-wav-files 
+# "/home/tensor/code/TTSDockerContainer/Vall-E-mount/input/20.wav" 
+# "/home/tensor/code/TTSDockerContainer/Vall-E-mount/input/21.wav" 
+# --whisper-folder-path "/home/tensor/code/storyteller/storyteller-ml/tts/VALL-E-X/whisper" 
+# --whisper-model medium 
+# --vocos-folder-path "/home/tensor/code/TTSDockerContainer/Vall-E-mount/models/vocos-encodec-24khz" 
+# --vallex-path "/home/tensor/code/TTSDockerContainer/Vall-E-mount/models/vallex-checkpoint.pt" 
+# --prompt-path "/home/tensor/code/storyteller/VALL-E-X-TTS-Container/Vall-E-mount/prompts" 
+# --prompt-name "test_prompt" 
+# --tmp-work-dir "/tmp"
diff --git a/tts/VALL-E-X/utils/prompt_making.py b/tts/VALL-E-X/utils/prompt_making.py
index d71a359..41f3848 100644
--- a/tts/VALL-E-X/utils/prompt_making.py
+++ b/tts/VALL-E-X/utils/prompt_making.py
@@ -84,6 +84,7 @@ def make_prompt(name:pathlib.Path, audio_prompt_path:pathlib.Path,audio_prompt_o
     else:
         save_path = os.path.join("./customs/", f"{name}.npz")
     np.savez(save_path, audio_tokens=audio_tokens, text_tokens=text_tokens, lang_code=lang2code[lang_pr])
+    print(f"Embedding Save Path {save_path}")
     logging.info(f"Successful. Prompt saved to {save_path}")
 
 def make_transcript(name, wav, sr, whisper_folder_path: pathlib.Path,transcript=None):