Let faster-whisper handle model path (#251)

* let faster-whisper handle model path

* fix quality ruff warnings

* remove useless tests

src/wordcab_transcribe/config.py

@@ -20,11 +20,9 @@
 """Configuration module of the Wordcab Transcribe."""
 
 from os import getenv
-from pathlib import Path
 from typing import Dict, List
 
 from dotenv import load_dotenv
-from faster_whisper.utils import _MODELS
 from loguru import logger
 from pydantic import field_validator
 from pydantic.dataclasses import dataclass
@@ -112,21 +110,6 @@ def api_prefix_must_not_be_none(cls, value: str):  # noqa: B902, N805
 
         return value
 
-    @field_validator("whisper_model")
-    def whisper_model_must_be_valid(cls, value: str):  # noqa: B902, N805
-        """Check that the model name is valid. It can be a local path or a model name."""
-        model_path = Path(value)
-
-        if model_path.exists() is False:
-            if value not in _MODELS.keys():
-                raise ValueError(
-                    f"{value} is not a valid model name. Choose one of"
-                    f" {_MODELS.keys()}.If you want to use a local model, please"
-                    " provide a valid path."
-                )
-
-        return value
-
     @field_validator("compute_type")
     def compute_type_must_be_valid(cls, value: str):  # noqa: B902, N805
         """Check that the model precision is valid."""

src/wordcab_transcribe/services/diarization/diarize_service.py

@@ -80,9 +80,7 @@ def __init__(
         else:
             self.default_segmentation_batch_size = 256
 
-        self.default_scale_dict = {
-            k: (w, s) for k, (w, s) in enumerate(zip(window_lengths, shift_lengths))
-        }
+        self.default_scale_dict = dict(enumerate(zip(window_lengths, shift_lengths)))
 
         for idx in device_index:
             _device = f"cuda:{idx}" if self.device == "cuda" else "cpu"
@@ -127,22 +125,18 @@ def __call__(
             segmentation_batch_size = self.default_segmentation_batch_size
             multiscale_weights = self.default_multiscale_weights
         elif audio_duration < 10800:
-            scale_dict = {
-                k: (w, s)
-                for k, (w, s) in enumerate(
+            scale_dict = dict(
+                enumerate(
                     zip(
                         [3.0, 2.5, 2.0, 1.5, 1.0],
                         self.default_shift_lengths,
                     )
                 )
-            }
+            )
             segmentation_batch_size = 64
             multiscale_weights = self.default_multiscale_weights
         else:
-            scale_dict = {
-                k: (w, s)
-                for k, (w, s) in enumerate(zip([3.0, 2.0, 1.0], [0.75, 0.5, 0.25]))
-            }
+            scale_dict = dict(enumerate(zip([3.0, 2.0, 1.0], [0.75, 0.5, 0.25])))
             segmentation_batch_size = 32
             multiscale_weights = [1.0, 1.0, 1.0]
 

tests/test_config.py

@@ -126,19 +126,6 @@ def test_general_parameters_validator(default_settings: dict) -> None:
         Settings(**wrong_api_prefix)
 
 
-def test_whisper_model_validator(default_settings: dict) -> None:
-    """Test whisper model validator."""
-    wrong_whisper_model = default_settings.copy()
-    wrong_whisper_model["whisper_model"] = "invalid_model_name"
-    with pytest.raises(ValueError):
-        Settings(**wrong_whisper_model)
-
-    wrong_whisper_model = default_settings.copy()
-    wrong_whisper_model["whisper_model"] = "/path/to/invalid_model"
-    with pytest.raises(ValueError):
-        Settings(**wrong_whisper_model)
-
-
 def test_compute_type_validator(default_settings: dict) -> None:
     """Test compute type validator."""
     default_settings["compute_type"] = "invalid_compute_type"