Merge pull request #310 from fireblade2534/master

Added some better saftey checks to captioned speech

api/src/routers/development.py

@@ -224,10 +224,13 @@ async def create_captioned_speech(
                                 ).decode("utf-8")
 
                                 # Add any chunks that may be in the acumulator into the return word_timestamps
-                                chunk_data.word_timestamps = (
+                                if chunk_data.word_timestamps is not None:
-                                    timestamp_acumulator + chunk_data.word_timestamps
+                                    chunk_data.word_timestamps = (
-                                )
+                                        timestamp_acumulator + chunk_data.word_timestamps
-                                timestamp_acumulator = []
+                                    )
+                                    timestamp_acumulator = []
+                                else:
+                                    chunk_data.word_timestamps = []
 
                                 yield CaptionedSpeechResponse(
                                     audio=base64_chunk,
@@ -272,7 +275,7 @@ async def create_captioned_speech(
                             )
 
                             # Add any chunks that may be in the acumulator into the return word_timestamps
-                            if chunk_data.word_timestamps != None:
+                            if chunk_data.word_timestamps is not None:
                                 chunk_data.word_timestamps = (
                                     timestamp_acumulator + chunk_data.word_timestamps
                                 )

dev/Test Phon.py

@@ -0,0 +1,23 @@
+import base64
+import json
+
+import pydub
+import requests
+
+def generate_audio_from_phonemes(phonemes: str, voice: str = "af_bella"):
+    """Generate audio from phonemes"""
+    response = requests.post(
+        "http://localhost:8880/dev/generate_from_phonemes",
+        json={"phonemes": phonemes, "voice": voice},
+        headers={"Accept": "audio/wav"}
+    )
+    if response.status_code != 200:
+        print(f"Error: {response.text}")
+        return None
+    return response.content
+
+
+
+
+with open(f"outputnostreammoney.wav", "wb") as f:
+    f.write(generate_audio_from_phonemes(r"mɪsəki ɪz ɐn ɪkspˌɛɹəmˈɛntᵊl ʤˈitəpˈi ˈɛnʤən dəzˈInd tə pˈWəɹ fjˈuʧəɹ vˈɜɹʒənz ʌv kəkˈɔɹO mˈɑdᵊlz."))