Auto-update: Thu Aug 8 06:18:49 PDT 2024

sijapi/routers/img.py

@@ -1,9 +1,5 @@
 '''
 Image generation module using StableDiffusion and similar models by way of ComfyUI.
-DEPENDS ON:
-  LLM module
-  COMFYUI_URL, COMFYUI_DIR, COMFYUI_OUTPUT_DIR, TS_SUBNET, TS_ADDRESS, DATA_DIR, IMG_CONFIG_DIR, IMG_DIR, IMG_WORKFLOWS_DIR, LOCAL_HOSTS
-*unimplemented.
 '''
 #routers/img.py
 

sijapi/routers/news.py

@@ -65,7 +65,7 @@ async def generate_summary(text: str) -> str:
     summary = await llm.summarize_text(text, "Summarize the provided text. Respond with the summary and nothing else.")
     return summary.replace('\n', ' ')
 
-async def handle_tts(bg_tasks: BackgroundTasks, article: Article, title: str, tts_mode: str, voice: str, summary: str) -> Optional[str]:
+async def handle_tts(bg_tasks: BackgroundTasks, article: Article, title: str, tts_mode: str, voice: str, summary: str, model: str = "eleven_turbo_v2") -> Optional[str]:
     if tts_mode in ["full", "content"]:
         tts_text = article.text
     elif tts_mode in ["summary", "excerpt"]:
@@ -79,12 +79,13 @@ async def handle_tts(bg_tasks: BackgroundTasks, article: Article, title: str, tt
             bg_tasks=bg_tasks,
             text=tts_text,
             voice=voice,
-            model="xtts",
+            model=model,
             podcast=True,
             title=audio_filename,
             output_dir=Path(OBSIDIAN_VAULT_DIR) / OBSIDIAN_RESOURCES_DIR
         )
         return f"![[{Path(audio_path).name}]]"
+
     except HTTPException as e:
         err(f"Failed to generate TTS: {str(e)}")
         return None
@@ -182,9 +183,6 @@ async def process_and_save_article(
                     # Save markdown file
                     await save_markdown_file(markdown_filename, markdown_content)
 
-                    # Add to daily note
-                    # await note.add_to_daily_note(relative_path)
-
                     return f"Successfully saved: {relative_path}"
                 
                 except Exception as e:

sijapi/routers/tts.py

@@ -147,6 +147,7 @@ async def generate_speech(
         
         if model == "eleven_turbo_v2":
             info("Using ElevenLabs.")
+            voice = determine_voice_id(voice)
             audio_file_path = await elevenlabs_tts(model, text, voice, title, output_dir)
         else:  # if model == "xtts":
             info("Using XTTS2")
@@ -182,6 +183,7 @@ async def get_model(voice: str = None, voice_file: UploadFile = None):
 async def determine_voice_id(voice_name: str) -> str:
     debug(f"Searching for voice id for {voice_name}")
     
+    # Todo: move this to tts.yaml
     hardcoded_voices = {
         "alloy": "E3A1KVbKoWSIKSZwSUsW",
         "echo": "b42GBisbu9r5m5n6pHF7",
@@ -192,7 +194,8 @@ async def determine_voice_id(voice_name: str) -> str:
         "Luna": "6TayTBKLMOsghG7jYuMX",
         "Sangye": "E7soeOyjpmuZFurvoxZ2",
         "Herzog": "KAX2Y6tTs0oDWq7zZXW7",
-        "Attenborough": "b42GBisbu9r5m5n6pHF7"
+        "Attenborough": "b42GBisbu9r5m5n6pHF7",
+        "Victoria": "7UBkHqZOtFRLq6cSMQQg"
     }
 
     if voice_name in hardcoded_voices:
@@ -212,6 +215,7 @@ async def determine_voice_id(voice_name: str) -> str:
                 for voice in voices_data:
                     if voice_name == voice["voice_id"] or voice_name == voice["name"]:
                         return voice["voice_id"]
+                    
         except Exception as e:
             err(f"Error determining voice ID: {str(e)}")