talk and talk-llama: get voices from api

ggerganov · Feb 14, 2024 · 53ad35c · 53ad35c
1 parent 7574276
commit 53ad35c
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 8 deletions.
diff --git a/examples/talk-llama/eleven-labs.py b/examples/talk-llama/eleven-labs.py
@@ -3,8 +3,7 @@
 
 parser = argparse.ArgumentParser(description="Generate the TTS")
 parser.add_argument("inputfile")
-parser.add_argument("-v", "--voice", type=int, default=2,
- choices=[0, 1, 2, 3, 4, 5, 6],
+parser.add_argument("-v", "--voice", type=int, default=21,
  help="Get a voice object by number")
 group = parser.add_mutually_exclusive_group()
 group.add_argument("-s", "--savefile", default="audio.mp3",
@@ -21,12 +20,14 @@
  print("elevenlabs library is not installed, you can install it to your enviroment using 'pip install elevenlabs'")
  sys.exit()
 
-from elevenlabs import generate, play, save
+from elevenlabs import voices, generate, play, save, DEFAULT_VOICE
 
 with open(args.inputfile) as f:
+ voicelist = [DEFAULT_VOICE]
+ voicelist += voices()[:]
  audio = generate(
  text=str(f.read()),
- voice=["Adam", "Antoni", "Arnold", "Bella", "Domi", "Elli", "Josh"][args.voice]
+ voice=voicelist[args.voice % len(voicelist)]
  )
  if args.play:
  play(audio)

diff --git a/examples/talk/eleven-labs.py b/examples/talk/eleven-labs.py
@@ -3,8 +3,7 @@
 
 parser = argparse.ArgumentParser(description="Generate the TTS")
 parser.add_argument("inputfile")
-parser.add_argument("-v", "--voice", type=int, default=2,
- choices=[0, 1, 2, 3, 4, 5, 6],
+parser.add_argument("-v", "--voice", type=int, default=21,
  help="Get a voice object by number")
 group = parser.add_mutually_exclusive_group()
 group.add_argument("-s", "--savefile", default="audio.mp3",
@@ -21,12 +20,14 @@
  print("elevenlabs library is not installed, you can install it to your enviroment using 'pip install elevenlabs'")
  sys.exit()
 
-from elevenlabs import generate, play, save
+from elevenlabs import voices, generate, play, save, DEFAULT_VOICE
 
 with open(args.inputfile) as f:
+ voicelist = [DEFAULT_VOICE]
+ voicelist += voices()[:]
  audio = generate(
  text=str(f.read()),
- voice=["Adam", "Antoni", "Arnold", "Bella", "Domi", "Elli", "Josh"][args.voice]
+ voice=voicelist[args.voice % len(voicelist)]
  )
  if args.play:
  play(audio)