0.2.0 rc5

docker-compose.yml

@@ -9,8 +9,8 @@ services:
     tty: true
     ports:
       - "8000:8000"
-    #volumes:
-    #  - .:/app/
+    volumes:
+      - .:/app/
     # Below can be removed if not using GPU
     runtime: nvidia
     deploy:

download_samples.sh

@@ -0,0 +1,7 @@
+#!/bin/sh
+for i in echo fable onyx nova shimmer; do
+	wget -q https://cdn.openai.com/API/docs/audio/$i.wav -O - | ffmpeg -loglevel error -i - -ar 22050 -ac 1 voices/$i.wav
+done
+
+# in testing alloy sounded REALY BAD after cloning. Save it anyways, but use another as the default.
+wget -q https://cdn.openai.com/API/docs/audio/alloy.wav -O - | ffmpeg -loglevel error -i - -ar 22050 -ac 1 voices/alloy0.wav

download_voices_tts-1-hd.sh

@@ -2,4 +2,4 @@
 export COQUI_TOS_AGREED=1
 model="tts_models/multilingual/multi-dataset/xtts_v2"
 python -c "from TTS.utils.manage import ModelManager; ModelManager().download_model('$model')"
-$(cd voices/ && ./download_samples.sh)
+./download_samples.sh

main.py

@@ -18,6 +18,7 @@ app = FastAPI()
 
 class xtts_wrapper():
     def __init__(self, model_name):
+        global args
         self.xtts = TTS(model_name=model_name, progress_bar=False).to(args.xtts_device)
 
     def tts(self, text, speaker_wav, speed):
@@ -57,6 +58,7 @@ class GenerateSpeechRequest(BaseModel):
 
 @app.post("/v1/audio/speech")
 async def generate_speech(request: GenerateSpeechRequest):
+    global xtts, args
     input_text = preprocess(request.input)
     model = request.model
     voice = request.voice
@@ -132,7 +134,7 @@ if __name__ == "__main__":
     parser.add_argument('--piper_cuda', action='store_true', default=False, help="Enable cuda for piper. Note: --cuda/onnxruntime-gpu is not working for me, but cpu is fast enough") 
     parser.add_argument('--xtts_device', action='store', default="cuda", help="Set the device for the xtts model. The special value of 'none' will use piper for all models.")
     parser.add_argument('--preload_xtts', action='store_true', default=False, help="Preload the xtts model. By default it's loaded on first use.")
-    parser.add_argument('-P', '--port', action='store', default=8000, help="Server tcp port")
+    parser.add_argument('-P', '--port', action='store', default=8000, type=int, help="Server tcp port")
     parser.add_argument('-H', '--host', action='store', default='localhost', help="Host to listen on, Ex. 0.0.0.0")
 
     args = parser.parse_args()

test_voices.sh

@@ -0,0 +1,38 @@
+#!/bin/sh
+
+curl -s http://localhost:8000/v1/audio/speech  -H "Content-Type: application/json"   -d "{
+    \"model\": \"tts-1\",
+    \"input\": \"I'm going to play you the original voice, followed by the piper voice and finally the X T T S version 2 voice\",
+    \"voice\": \"echo\",
+    \"speed\": 1.0
+  }" | mpv --really-quiet -
+
+for voice in  alloy echo fable onyx nova shimmer ; do
+
+echo $voice
+
+curl -s http://localhost:8000/v1/audio/speech  -H "Content-Type: application/json"   -d "{
+    \"model\": \"tts-1\",
+    \"input\": \"original\",
+    \"voice\": \"echo\",
+    \"speed\": 1.0
+  }" | mpv --really-quiet -
+
+wget -q https://cdn.openai.com/API/docs/audio/$voice.wav -O - | mpv --really-quiet -
+
+curl -s http://localhost:8000/v1/audio/speech  -H "Content-Type: application/json"   -d "{
+    \"model\": \"tts-1\",
+    \"input\": \"The quick brown fox jumped over the lazy dog. This voice is called $voice, how do you like this voice?\",
+    \"voice\": \"$voice\",
+    \"speed\": 1.0
+  }" | mpv --really-quiet -
+
+curl -s http://localhost:8000/v1/audio/speech  -H "Content-Type: application/json"   -d "{
+    \"model\": \"tts-1-hd\",
+    \"input\": \"The quick brown fox jumped over the lazy dog. This HD voice is called $voice, how do you like this voice?\",
+    \"voice\": \"$voice\",
+    \"speed\": 1.0
+  }" | mpv --really-quiet -
+
+done
+

voice_to_speaker.yaml

@@ -32,13 +32,13 @@ tts-1-hd:
     speaker: voices/fable.wav
   onyx:
     model: tts_models/multilingual/multi-dataset/xtts_v2
-    speaker: voices/voices/onyx.wav
+    speaker: voices/onyx.wav
   nova:
     model: tts_models/multilingual/multi-dataset/xtts_v2
-    speaker: voices/voices/nova.wav
+    speaker: voices/nova.wav
   shimmer:
     model: tts_models/multilingual/multi-dataset/xtts_v2
-    speaker: voices/voices/shimmer.wav
+    speaker: voices/shimmer.wav
   me:
     model: tts_models/multilingual/multi-dataset/xtts_v2
-    speaker: voices/voices/me.wav # this could be you
+    speaker: voices/me.wav # this could be you

voices/download_samples.sh

@@ -1,7 +0,0 @@
-#!/bin/sh
-wget -q https://cdn.openai.com/API/docs/audio/alloy.wav
-wget -q https://cdn.openai.com/API/docs/audio/echo.wav
-wget -q https://cdn.openai.com/API/docs/audio/fable.wav
-wget -q https://cdn.openai.com/API/docs/audio/onyx.wav
-wget -q https://cdn.openai.com/API/docs/audio/nova.wav
-wget -q https://cdn.openai.com/API/docs/audio/shimmer.wav