Merge branch 'develop'

api/commands.py

@@ -21,18 +21,14 @@ class SaveSecretCommand(WebSocketCommandModel):
     command: Literal["save_secret"] = "save_secret"
     secret_name: str
     secret_value: str
-    show_message: bool = True
+    show_message: Optional[bool] = True
 
 
 class RecordKeyboardActionsCommand(WebSocketCommandModel):
     command: Literal["record_keyboard_actions"] = "record_keyboard_actions"
     recording_type: KeyboardRecordingType = KeyboardRecordingType.SINGLE
 
 
-class RecordMouseActionsCommand(WebSocketCommandModel):
-    command: Literal["record_mouse_actions"] = "record_mouse_actions"
-
-
 class StopRecordingCommand(WebSocketCommandModel):
     command: Literal["stop_recording"] = "stop_recording"
 
@@ -44,7 +40,7 @@ class LogCommand(WebSocketCommandModel):
     command: Literal["log"] = "log"
     text: str
     log_type: LogType
-    source_name: str = None
+    source_name: Optional[str] = None
     source: LogSource = "system"
     tag: Optional[CommandTag] = None
 

api/enums.py

@@ -68,9 +68,27 @@ class OpenAiModel(Enum):
     GPT_4_TURBO_PREVIEW = "gpt-4-turbo-preview"
 
 
+class MistralModel(Enum):
+    MISTRAL_7B = "open-mistral-7b"
+    OPEN_MIXTRAL_8X7B = "open-mixtral-8x7b"
+    MISTRAL_SMALL = "mistral-small-latest"
+    MISTRAL_MEDIUM = "mistral-medium-latest"
+    MISTRAL_LARGE = "mistral-large-latest"
+
+
+class LlamaModel(Enum):
+    LLAMA3_8B = "llama3-8b"
+    LLAMA3_8B_8192 = "llama3-8b-8192"
+    LLAMA3_70B = "llama3-70b"
+    LLAMA3_70B_8192 = "llama3-70b-8192"
+
+
 class WingmanProAzureDeployment(Enum):
     GPT_35_TURBO = "gpt-35-turbo"
     GPT_4_TURBO = "gpt-4-turbo"
+    MISTRAL_LARGE = "mistral-large-latest"
+    LLAMA3_8B = "llama3-8b"
+    LLAMA3_70B = "llama3-70b"
 
 
 class OpenAiTtsVoice(Enum):
@@ -117,12 +135,16 @@ class VoiceActivationSttProvider(Enum):
 
 class ConversationProvider(Enum):
     OPENAI = "openai"
+    MISTRAL = "mistral"
+    LLAMA = "llama"
     AZURE = "azure"
     WINGMAN_PRO = "wingman_pro"
 
 
 class SummarizeProvider(Enum):
     OPENAI = "openai"
+    MISTRAL = "mistral"
+    LLAMA = "llama"
     AZURE = "azure"
     WINGMAN_PRO = "wingman_pro"
 
@@ -197,6 +219,14 @@ class OpenAiModelEnumModel(BaseEnumModel):
     model: OpenAiModel
 
 
+class MistralModelEnumModel(BaseEnumModel):
+    model: MistralModel
+
+
+class LlamaModelEnumModel(BaseEnumModel):
+    model: LlamaModel
+
+
 class WingmanProAzureDeploymentEnumModel(BaseEnumModel):
     deployment_name: WingmanProAzureDeployment
 
@@ -260,6 +290,8 @@ class WingmanProTtsProviderModel(BaseEnumModel):
     "ElevenlabsModel": ElevenlabsModelEnumModel,
     "TtsVoiceGender": TtsVoiceGenderEnumModel,
     "OpenAiModel": OpenAiModelEnumModel,
+    "MistralModel": MistralModelEnumModel,
+    "LLamaModel": LlamaModelEnumModel,
     "WingmanProAzureDeployment": WingmanProAzureDeploymentEnumModel,
     "OpenAiTtsVoice": OpenAiTtsVoiceEnumModel,
     "SoundEffect": SoundEffectEnumModel,

api/interface.py

@@ -9,6 +9,8 @@
     AzureApiVersion,
     AzureRegion,
     ConversationProvider,
+    LlamaModel,
+    MistralModel,
     TtsVoiceGender,
     ElevenlabsModel,
     OpenAiModel,
@@ -101,9 +103,14 @@ class AudioDevice(BaseModel):
 # CONFIG MODELS
 
 
+class AudioDeviceSettings(BaseModel):
+    hostapi: Optional[int] = 0
+    name: str
+
+
 class AudioSettings(BaseModel):
-    input: Optional[int] = None
-    output: Optional[int] = None
+    input: Optional[int | AudioDeviceSettings] = None
+    output: Optional[int | AudioDeviceSettings] = None
 
 
 class WhispercppAutostartSettingsConfig(BaseModel):
@@ -269,6 +276,18 @@ class OpenAiConfig(BaseModel):
     """If you have an organization key, you can set it here."""
 
 
+class MistralConfig(BaseModel):
+    conversation_model: MistralModel
+    summarize_model: MistralModel
+    endpoint: str
+
+
+class LlamaConfig(BaseModel):
+    conversation_model: LlamaModel
+    summarize_model: LlamaModel
+    endpoint: str
+
+
 class WingmanProConfig(BaseModel):
     stt_provider: WingmanProSttProvider
     tts_provider: WingmanProTtsProvider
@@ -345,6 +364,7 @@ class CommandKeyboardConfig(BaseModel):
     release: Optional[bool] = None
     """Whether to release the key. Optional."""
 
+
 class CommandMouseConfig(BaseModel):
     button: Optional[str] = None
     """The mouse button to press. Optional."""
@@ -410,6 +430,8 @@ class NestedConfig(BaseModel):
     sound: SoundConfig
     features: FeaturesConfig
     openai: OpenAiConfig
+    mistral: MistralConfig
+    llama: LlamaConfig
     edge_tts: EdgeTtsConfig
     elevenlabs: ElevenlabsConfig
     azure: AzureConfig

configs/templates/_Star Citizen/Computer.template.yaml

@@ -412,34 +412,3 @@ commands:
       - Launch Missile
       - Fire Missile
   # ───────────────────────────────────────────
-  - name: MoveTurretLeft
-    actions:
-      - mouse:
-          move: [-100, 0]
-    instant_activation:
-      - Turret left
-  # ───────────────────────────────────────────
-  - name: MoveTurretRight
-    actions:
-      - mouse:
-          move: [100, 0]
-    instant_activation:
-      - Turret right
-  # ───────────────────────────────────────────
-  - name: MoveTurretUp
-    actions:
-      - mouse:
-          button: left
-          hold: 1
-          move: [0, -100]
-    instant_activation:
-      - Turret up
-  # ───────────────────────────────────────────
-  - name: MoveTurretDown
-    actions:
-      - mouse:
-          button: left
-          hold: 1
-          move: [0, 100]
-    instant_activation:
-      - Turret down

configs/templates/defaults.yaml

@@ -81,6 +81,16 @@ openai:
   #base_url: https://api.openai.com # or the localhost address of your local LLM etc.
   #organization: whatever # If you have an organization key, you can set it here.
 
+mistral:
+  conversation_model: mistral-large-latest
+  summarize_model: mistral-medium-latest
+  endpoint: https://api.mistral.ai/v1
+
+llama:
+  conversation_model: llama3-70b
+  summarize_model: llama3-8b
+  endpoint: https://api.groq.com/openai/v1
+
 # ────────────────────────────────── EDGE TTS ────────────────────────────────────
 # EdgeTTS is free and faster than the default OpenAI TTSbut it's not as good in terms of quality.
 # Only used if features > tts_provider is set to 'edge_tts' above.

providers/open_ai.py

@@ -4,7 +4,14 @@
 from typing import Literal
 from openai import OpenAI, APIStatusError, AzureOpenAI
 import azure.cognitiveservices.speech as speechsdk
-from api.enums import AzureRegion, LogType, OpenAiModel, OpenAiTtsVoice
+from api.enums import (
+    AzureRegion,
+    LlamaModel,
+    LogType,
+    MistralModel,
+    OpenAiModel,
+    OpenAiTtsVoice,
+)
 from api.interface import (
     AzureConfig,
     AzureInstanceConfig,
@@ -69,7 +76,7 @@ def _perform_ask(
         self,
         client: OpenAI | AzureOpenAI,
         messages: list[dict[str, str]],
-        model: OpenAiModel,
+        model: OpenAiModel | MistralModel | LlamaModel,
         stream: bool,
         tools: list[dict[str, any]],
     ):
@@ -133,7 +140,7 @@ def transcribe(self, filename: str, model: str = "whisper-1"):
     def ask(
         self,
         messages: list[dict[str, str]],
-        model: OpenAiModel,
+        model: OpenAiModel | MistralModel | LlamaModel,
         stream: bool = False,
         tools: list[dict[str, any]] = None,
     ):

requirements.txt

@@ -1,21 +1,22 @@
-azure-cognitiveservices-speech==1.36.0
+azure-cognitiveservices-speech==1.37.0
 edge-tts==6.1.10
-elevenlabslib~=0.21.0
-fastapi~=0.110.0
+elevenlabslib~=0.22.1
+fastapi~=0.110.1
 numpy~=1.26.4
-openai~=1.14.3
+openai~=1.23.2
 packaging~=24.0
 pedalboard~=0.9.3
 platformdirs~=4.2.0
 pyaudio~=0.2.14
-pydantic~=2.6.4
+pydantic~=2.7.0
 pydirectinput-rgx==2.1.1
-pyinstaller==6.5.0
+pyinstaller==6.6.0
+python-multipart==0.0.9
 PyYAML~=6.0.1
 requests~=2.31.0
-scipy~=1.12.0
+scipy~=1.13.0
 sounddevice~=0.4.6
 soundfile~=0.12.1
-SpeechRecognition~=3.10.1
-typing_extensions~=4.10.0
+SpeechRecognition~=3.10.3
+typing_extensions~=4.11.0
 uvicorn~=0.29.0