From 97cfbfee1d11363ee1d54b0b704d8361d49c6f11 Mon Sep 17 00:00:00 2001
From: Lewis Wolfgang <lewis@daily.co>
Date: Mon, 5 Aug 2024 10:01:27 -0400
Subject: [PATCH] Install silero via pip

---
 pyproject.toml            |  2 +-
 src/pipecat/vad/silero.py | 19 ++-----------------
 2 files changed, 3 insertions(+), 18 deletions(-)

diff --git a/pyproject.toml b/pyproject.toml
index 33a72a041..3c94aadf4 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -50,7 +50,7 @@ moondream = [ "einops~=0.8.0", "timm~=0.9.16", "transformers~=4.40.2" ]
 openai = [ "openai~=1.35.0" ]
 openpipe = [ "openpipe~=4.18.0" ]
 playht = [ "pyht~=0.0.28" ]
-silero = [ "torch~=2.3.1", "torchaudio~=2.3.1" ]
+silero = [ "silero-vad~=5.1" ]
 websocket = [ "websockets~=12.0", "fastapi~=0.111.0" ]
 whisper = [ "faster-whisper~=1.0.3" ]
 xtts = [ "resampy~=0.4.3" ]
diff --git a/src/pipecat/vad/silero.py b/src/pipecat/vad/silero.py
index 4fdf5c767..b801b4545 100644
--- a/src/pipecat/vad/silero.py
+++ b/src/pipecat/vad/silero.py
@@ -15,6 +15,7 @@
 from loguru import logger
 
 try:
+    from silero_vad import load_silero_vad
     import torch
     # We don't use torchaudio here, but we need to try importing it because
     # Silero uses it.
@@ -37,10 +38,6 @@ def __init__(
             self,
             *,
             sample_rate: int = 16000,
-            version: str = "v5.0",
-            force_reload: bool = False,
-            skip_validation: bool = True,
-            trust_repo: bool = True,
             params: VADParams = VADParams()):
         super().__init__(sample_rate=sample_rate, num_channels=1, params=params)
 
@@ -49,11 +46,7 @@ def __init__(
 
         logger.debug("Loading Silero VAD model...")
 
-        (self._model, _) = torch.hub.load(repo_or_dir=f"snakers4/silero-vad:{version}",
-                                          model="silero_vad",
-                                          force_reload=force_reload,
-                                          skip_validation=skip_validation,
-                                          trust_repo=trust_repo)
+        self._model = load_silero_vad()
 
         self._last_reset_time = 0
 
@@ -94,20 +87,12 @@ def __init__(
             self,
             *,
             sample_rate: int = 16000,
-            version: str = "v5.0",
-            force_reload: bool = False,
-            skip_validation: bool = True,
-            trust_repo: bool = True,
             vad_params: VADParams = VADParams(),
             audio_passthrough: bool = False):
         super().__init__()
 
         self._vad_analyzer = SileroVADAnalyzer(
             sample_rate=sample_rate,
-            version=version,
-            force_reload=force_reload,
-            skip_validation=skip_validation,
-            trust_repo=trust_repo,
             params=vad_params)
         self._audio_passthrough = audio_passthrough