change transcribe mechanism when uploading audio

Dockerfile

@@ -3,6 +3,11 @@
 
 FROM python:3.9
 
+# Set proper Python threading configuration
+ENV OMP_NUM_THREADS=4
+ENV NUMEXPR_NUM_THREADS=4
+ENV MKL_NUM_THREADS=4
+
 RUN apt update
 RUN apt --yes install ffmpeg
 
@@ -17,4 +22,4 @@ RUN pip install --no-cache-dir --upgrade -r requirements.txt
 
 COPY --chown=user . /app
 # CMD ["gunicorn", "-b", "0.0.0.0:7860", "app:app"]
-CMD ["gunicorn", "--timeout", "60", "-b", "0.0.0.0:7860", "app:app"]
+CMD ["gunicorn", "--timeout", "120", "--workers", "2", "-b", "0.0.0.0:7860", "app:app"]

app.py

@@ -49,32 +49,33 @@ model = MultiClassModel.load_from_checkpoint(
 )
 model.eval()
 
+# === INITIAL SETUP: Whisper Pipeline ===
+# https://huggingface.co/openai/whisper-large-v3
+device = "cuda:0" if torch.cuda.is_available() else "cpu"
+torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
+
+model_id = "openai/whisper-large-v3"
+
+whisper_model = AutoModelForSpeechSeq2Seq.from_pretrained(
+    model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True
+)
+whisper_model.to(device)
+
+processor = AutoProcessor.from_pretrained(model_id)
+
+pipe = pipeline(
+    "automatic-speech-recognition",
+    model=whisper_model,
+    tokenizer=processor.tokenizer,
+    feature_extractor=processor.feature_extractor,
+    chunk_length_s=10,
+    batch_size=4,  # batch size for inference - set based on your device
+    torch_dtype=torch_dtype,
+    device=device,
+)
 
-def whisper_api(temp_audio_path):
-    # https://huggingface.co/openai/whisper-large-v3
-    device = "cuda:0" if torch.cuda.is_available() else "cpu"
-    torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
-
-    model_id = "openai/whisper-large-v3"
-
-    model = AutoModelForSpeechSeq2Seq.from_pretrained(
-        model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True
-    )
-    model.to(device)
-
-    processor = AutoProcessor.from_pretrained(model_id)
-
-    pipe = pipeline(
-        "automatic-speech-recognition",
-        model=model,
-        tokenizer=processor.tokenizer,
-        feature_extractor=processor.feature_extractor,
-        chunk_length_s=10,
-        batch_size=4,  # batch size for inference - set based on your device
-        torch_dtype=torch_dtype,
-        device=device,
-    )
 
+def whisper_api(temp_audio_path):
     result = pipe(temp_audio_path, return_timestamps=False, generate_kwargs={"language": "indonesian"})
     print(result["text"])
     return result