more fixes

cog.yaml

@@ -18,7 +18,7 @@ build:
 
   # a list of packages in the format <package-name>==<version>
   python_packages:
-     - "faster-whisper==0.7.0"
+     - "faster-whisper==0.7.1"
      - "transformers==4.25.1"
      - "accelerate==0.15.0"
      - "ffmpeg-python==0.2.0"

predict.py

@@ -32,7 +32,7 @@ def setup(self):
         model_name = "large-v2"
         self.model = WhisperModel(
             model_name,
-            device=("cuda" if torch.cuda.is_available() else "cpu"),
+            device="cuda" if torch.cuda.is_available() else "cpu",
             compute_type="float16")
         self.embedding_model = PretrainedSpeakerEmbedding(
             "speechbrain/spkrec-ecapa-voxceleb",
@@ -68,7 +68,7 @@ def predict(
             raise RuntimeError("Provide either file_string or file_url")
         """ filepath = ''
         file_start, file_ending = os.path.splitext(f'{filename}')"""
-        ts = time.time()
+        ts = time.time_ns()
         filename = f'{ts}-recording'
         file_extension = '.mp3'
 
@@ -148,7 +148,6 @@ def speech_to_text(self,
         segments, _ = self.model.transcribe(audio_file_wav, **options)
         segments = list(segments)
         print("done with whisper")
-        print(segments)
         segments = [{
             'start':
             int(round(s.start + offset_seconds)),