fixed channels problem

configs/config.yaml

@@ -1,5 +1,5 @@
 # =================== Lord Of Large Language Multimodal Systems Configuration file =========================== 
-version: 99
+version: 100
 binding_name: null
 model_name: null
 model_variant: null
@@ -89,7 +89,11 @@ active_ttm_service: "None" # musicgen (offline)
 # -------------------- Services --------------------------
 
 # ***************** STT *****************
-# STT service 
+stt_input_device: null
+
+
+
+# ASR STT service 
 asr_enable: false
 asr_base_url: http://localhost:9000
 
@@ -104,6 +108,8 @@ whisper_model: base
 
 
 # ***************** TTS *****************
+tts_output_device: null
+
 # Voice service
 xtts_enable: false
 xtts_base_url: http://localhost:8020

lollms/configs/config.yaml

@@ -1,5 +1,5 @@
 # =================== Lord Of Large Language Multimodal Systems Configuration file =========================== 
-version: 99
+version: 100
 binding_name: null
 model_name: null
 model_variant: null
@@ -89,7 +89,11 @@ active_ttm_service: "None" # musicgen (offline)
 # -------------------- Services --------------------------
 
 # ***************** STT *****************
-# STT service 
+stt_input_device: null
+
+
+
+# ASR STT service 
 asr_enable: false
 asr_base_url: http://localhost:9000
 
@@ -104,6 +108,8 @@ whisper_model: base
 
 
 # ***************** TTS *****************
+tts_output_device: null
+
 # Voice service
 xtts_enable: false
 xtts_base_url: http://localhost:8020

lollms/media.py

@@ -109,7 +109,8 @@ class RTCom:
                         channels=1, 
                         buffer_size=10, 
                         model="small.en", 
-                        snd_device=None, 
+                        snd_input_device=None,
+                        snd_output_device=None,
                         logs_folder="logs", 
                         voice=None, 
                         block_while_talking=True, 
@@ -138,11 +139,15 @@ class RTCom:
         self.block_while_talking = block_while_talking
         self.image_shot = None
 
-        if snd_device is None:
+        if snd_input_device is None:
             devices = sd.query_devices()
-            snd_device = [device['name'] for device in devices][0]
+            snd_input_device = [device['name'] for device in devices if device['type'] == 'input'][0]
+        if snd_output_device is None:
+            devices = sd.query_devices()
+            snd_output_device = [device['name'] for device in devices if device['type'] == 'output'][0]
 
-        self.snd_device = snd_device
+        self.snd_input_device = snd_input_device
+        self.snd_output_device = snd_output_device
         self.logs_folder = logs_folder
 
         self.frames = []
@@ -206,8 +211,7 @@ class RTCom:
         ASCIIColors.green("<<RTCOM off>>")
 
     def _record(self):
-        sd.default.device = self.snd_device
-        with sd.InputStream(channels=self.channels, samplerate=self.rate, callback=self.callback, dtype='int16'):
+        with sd.InputStream(channels=self.channels, device=self.snd_input_device, samplerate=self.rate, callback=self.callback, dtype='int16'):
             while not self.stop_flag:
                 time.sleep(0.1)
 

lollms/server/configs/config.yaml

@@ -1,5 +1,5 @@
 # =================== Lord Of Large Language Multimodal Systems Configuration file =========================== 
-version: 95
+version: 100
 binding_name: null
 model_name: null
 model_variant: null
@@ -80,10 +80,36 @@ auto_show_browser: true
 # copy to clipboard 
 copy_to_clipboard_add_all_details: false
 
-# STT service 
+# -------------------- Services global configurations --------------------------
+# Select the active test to speach, text to image and speach to text services
+active_tts_service: "None" # xtts (offline), openai_tts (API key required)
+active_tti_service: "None" # autosd (offline), dall-e (online)
+active_stt_service: "None" # whisper (offline), asr (offline or online), openai_whiosper (API key required)
+active_ttm_service: "None" # musicgen (offline)
+# -------------------- Services --------------------------
+
+# ***************** STT *****************
+stt_input_device: null
+
+
+
+# ASR STT service 
 asr_enable: false
 asr_base_url: http://localhost:9000
 
+# openai_whisper configuration
+openai_whisper_key: ""
+openai_whisper_model: "whisper-1"
+
+
+# whisper configuration
+whisper_activate: false
+whisper_model: base
+
+
+# ***************** TTS *****************
+tts_output_device: null
+
 # Voice service
 xtts_enable: false
 xtts_base_url: http://localhost:8020
@@ -101,10 +127,24 @@ xtts_top_p: 0.85
 xtts_speed: 1
 xtts_enable_text_splitting: true
 
+# openai_whisper configuration
+openai_tts_key: ""
+openai_tts_model: "tts-1"
+openai_tts_voice: "alloy"
+
+# ***************** TTI *****************
+
 # Image generation service
 enable_sd_service: false
 sd_base_url: http://localhost:7860
 
+# Dall e service key
+dall_e_key: ""
+dall_e_generation_engine: "dall-e-3"
+
+# Midjourney service key
+midjourney_key: ""
+
 # Image generation service comfyui
 enable_comfyui_service: false
 comfyui_base_url: http://127.0.0.1:8188/
@@ -113,6 +153,8 @@ comfyui_base_url: http://127.0.0.1:8188/
 enable_motion_ctrl_service: false
 motion_ctrl_base_url: http://localhost:7861
 
+# ***************** TTT *****************
+
 # ollama service
 enable_ollama_service: false
 ollama_base_url: http://localhost:11434
@@ -197,6 +239,3 @@ show_code_of_conduct: true
 activate_audio_infos: true
 
 
-# whisper configuration
-whisper_activate: false
-whisper_model: base

lollms/server/endpoints/lollms_binding_files_server.py

@@ -19,6 +19,7 @@ from ascii_colors import ASCIIColors
 from lollms.utilities import load_config, trace_exception, gc
 from pathlib import Path
 from typing import List
+from lollms.security import sanitize_svg
 import os
 import re
 
@@ -274,6 +275,19 @@ async def serve_discussions(path: str):
     if not Path(file_path).exists():
         raise HTTPException(status_code=404, detail="File not found")
 
+    # Check if the file is an SVG
+    if file_path.suffix.lower() == '.svg':
+        with open(file_path, 'r', encoding='utf-8') as file:
+            svg_content = file.read()
+        sanitized_svg_content = sanitize_svg(svg_content)
+        
+        # Save the sanitized SVG content to a temporary file
+        temp_svg_path = file_path.with_suffix('.sanitized.svg')
+        with open(temp_svg_path, 'w', encoding='utf-8') as file:
+            file.write(sanitized_svg_content)
+        
+        return FileResponse(str(temp_svg_path))
+
     return FileResponse(str(file_path))
 
 

lollms/server/endpoints/lollms_tts.py

@@ -304,3 +304,13 @@ def tts_is_ready():
         if lollmsElfServer.tts.ready:
             return {"status":True}
     return {"status":False}
+
+
+@router.get("/get_snd_input_devices")
+def get_snd_input_devices():
+    lollmsElfServer.stt.get_devices()
+
+@router.get("/get_snd_output_devices")
+def get_snd_output_devices():
+    lollmsElfServer.tts.get_devices()
+

lollms/stt.py

@@ -8,7 +8,24 @@ Author: ParisNeo, a computer geek passionate about AI
 """
 
 from lollms.app import LollmsApplication
+from lollms.utilities import PackageManager
 from pathlib import Path
+from ascii_colors import ASCIIColors
+
+try:
+    if not PackageManager.check_package_installed("sounddevice"):
+        # os.system("sudo apt-get install portaudio19-dev")
+        PackageManager.install_package("sounddevice")
+        PackageManager.install_package("wave")
+except:
+    # os.system("sudo apt-get install portaudio19-dev -y")
+    PackageManager.install_package("sounddevice")
+    PackageManager.install_package("wave")
+try:
+    import sounddevice as sd
+    import wave
+except:
+    ASCIIColors.error("Couldn't load sound tools")
 
 class LollmsSTT:
     """
@@ -96,3 +113,11 @@ class LollmsSTT:
             LollmsSTT: The LollmsSTT class.
         """
         return LollmsSTT
+
+    def get_devices(self):
+        devices =  sd.query_devices()
+        print(devices)
+        return {
+            "status": True,
+            "device_names": [device['name'] for device in devices if device["max_input_channels"]>0]
+        }

lollms/tts.py

@@ -7,8 +7,23 @@ This module is part of the Lollms library, designed to provide Text-to-Speech (T
 Author: ParisNeo, a computer geek passionate about AI
 """
 from lollms.app import LollmsApplication
+from lollms.utilities import PackageManager
 from pathlib import Path
-
+from ascii_colors import ASCIIColors
+try:
+    if not PackageManager.check_package_installed("sounddevice"):
+        # os.system("sudo apt-get install portaudio19-dev")
+        PackageManager.install_package("sounddevice")
+        PackageManager.install_package("wave")
+except:
+    # os.system("sudo apt-get install portaudio19-dev -y")
+    PackageManager.install_package("sounddevice")
+    PackageManager.install_package("wave")
+try:
+    import sounddevice as sd
+    import wave
+except:
+    ASCIIColors.error("Couldn't load sound tools")
 class LollmsTTS:
     """
     LollmsTTS is a base class for implementing Text-to-Speech (TTS) functionalities within the LollmsApplication.
@@ -119,4 +134,12 @@ class LollmsTTS:
         Returns:
             list: A list of available voices.
         """
-        return self.voices
+        return self.voices
+    
+    def get_devices(self):
+        devices =  sd.query_devices()
+
+        return {
+            "status": True,
+            "device_names": [device['name'] for device in devices if device["max_output_channels"]>0]
+        }