Fix model download

server/gpu/modal_deployments/reflector_diarizer.py

@@ -30,16 +30,18 @@ def migrate_cache_llm():
 
 def download_pyannote_audio():
     from pyannote.audio import Pipeline
+
     Pipeline.from_pretrained(
-        "pyannote/speaker-diarization-3.0",
+        PYANNOTE_MODEL_NAME,
         cache_dir=MODEL_DIR,
+        use_auth_token=os.environ["HF_TOKEN"],
     )
 
 
 diarizer_image = (
     Image.debian_slim(python_version="3.10.8")
     .pip_install(
-        "pyannote.audio",
+        "pyannote.audio==3.1.0",
         "requests",
         "onnx",
         "torchaudio",
@@ -50,10 +52,12 @@ diarizer_image = (
         "protobuf",
         "numpy",
         "huggingface_hub",
-        "hf-transfer"
+        "hf-transfer",
+    )
+    .run_function(
+        download_pyannote_audio, secrets=[Secret.from_name("my-huggingface-secret")]
     )
     .run_function(migrate_cache_llm)
-    .run_function(download_pyannote_audio)
     .env(
         {
             "LD_LIBRARY_PATH": (
@@ -64,6 +68,7 @@ diarizer_image = (
     )
 )
 
+
 @app.cls(
     gpu=modal.gpu.A100(size="40GB"),
     timeout=60 * 30,
@@ -80,18 +85,12 @@ class Diarizer:
         self.use_gpu = torch.cuda.is_available()
         self.device = "cuda" if self.use_gpu else "cpu"
         self.diarization_pipeline = Pipeline.from_pretrained(
-            "pyannote/speaker-diarization-3.0",
-            cache_dir=MODEL_DIR
+            PYANNOTE_MODEL_NAME, cache_dir=MODEL_DIR
         )
         self.diarization_pipeline.to(torch.device(self.device))
 
     @method()
-    def diarize(
-            self,
-            audio_data: str,
-            audio_suffix: str,
-            timestamp: float
-    ):
+    def diarize(self, audio_data: str, audio_suffix: str, timestamp: float):
         import tempfile
 
         import torchaudio
@@ -101,21 +100,24 @@ class Diarizer:
 
             print("Diarizing audio")
             waveform, sample_rate = torchaudio.load(fp.name)
-            diarization = self.diarization_pipeline({"waveform": waveform, "sample_rate": sample_rate})
+            diarization = self.diarization_pipeline(
+                {"waveform": waveform, "sample_rate": sample_rate}
+            )
 
             words = []
-            for diarization_segment, _, speaker in diarization.itertracks(yield_label=True):
+            for diarization_segment, _, speaker in diarization.itertracks(
+                yield_label=True
+            ):
                 words.append(
                     {
                         "start": round(timestamp + diarization_segment.start, 3),
                         "end": round(timestamp + diarization_segment.end, 3),
-                        "speaker": int(speaker[-2:])
+                        "speaker": int(speaker[-2:]),
                     }
                 )
             print("Diarization complete")
-            return {
-                "diarization": words
-            }
+            return {"diarization": words}
+
 
 # -------------------------------------------------------------------
 # Web API
@@ -129,7 +131,7 @@ class Diarizer:
     secrets=[
         Secret.from_name("reflector-gpu"),
     ],
-    image=diarizer_image
+    image=diarizer_image,
 )
 @asgi_app()
 def web():
@@ -157,16 +159,17 @@ def web():
         if response.status_code == 404:
             raise HTTPException(
                 status_code=response.status_code,
-                detail="The audio file does not exist."
+                detail="The audio file does not exist.",
             )
 
     class DiarizationResponse(BaseModel):
         result: dict
 
-    @app.post("/diarize", dependencies=[Depends(apikey_auth), Depends(validate_audio_file)])
+    @app.post(
+        "/diarize", dependencies=[Depends(apikey_auth), Depends(validate_audio_file)]
+    )
     def diarize(
-            audio_file_url: str,
-            timestamp: float = 0.0
+        audio_file_url: str, timestamp: float = 0.0
     ) -> HTTPException | DiarizationResponse:
         # Currently the uploaded files are in mp3 format
         audio_suffix = "mp3"
@@ -176,9 +179,7 @@ def web():
         print("Audio file downloaded successfully")
 
         func = diarizerstub.diarize.spawn(
-            audio_data=response.content,
-            audio_suffix=audio_suffix,
-            timestamp=timestamp
+            audio_data=response.content, audio_suffix=audio_suffix, timestamp=timestamp
         )
         result = func.get()
         return result