Fix typos, use idiomatic PyTorch API, and correct type hints

whisper/__init__.py

@@ -103,7 +103,7 @@ def available_models() -> List[str]:
 def load_model(
     name: str,
     device: Optional[Union[str, torch.device]] = None,
-    download_root: str = None,
+    download_root: Optional[str] = None,
     in_memory: bool = False,
 ) -> Whisper:
     """

whisper/timing.py

@@ -27,9 +27,9 @@ def median_filter(x: torch.Tensor, filter_width: int):
         # `F.pad` does not support 1D or 2D inputs for reflect padding but supports 3D and 4D
         x = x[None, None, :]
 
-    assert (
-        filter_width > 0 and filter_width % 2 == 1
-    ), "`filter_width` should be an odd number"
+    assert filter_width > 0 and filter_width % 2 == 1, (
+        "`filter_width` should be an odd number"
+    )
 
     result = None
     x = F.pad(x, (filter_width // 2, filter_width // 2, 0, 0), mode="reflect")
@@ -211,7 +211,7 @@ def find_alignment(
     weights = (weights - mean) / std
     weights = median_filter(weights, medfilt_width)
 
-    matrix = weights.mean(axis=0)
+    matrix = weights.mean(dim=0)
     matrix = matrix[len(tokenizer.sot_sequence) : -1]
     text_indices, time_indices = dtw(-matrix)
 

whisper/transcribe.py

@@ -561,7 +561,7 @@ def valid_model_name(name):
     parser.add_argument("--max_line_width", type=optional_int, default=None, help="(requires --word_timestamps True) the maximum number of characters in a line before breaking the line")
     parser.add_argument("--max_line_count", type=optional_int, default=None, help="(requires --word_timestamps True) the maximum number of lines in a segment")
     parser.add_argument("--max_words_per_line", type=optional_int, default=None, help="(requires --word_timestamps True, no effect with --max_line_width) the maximum number of words in a segment")
-    parser.add_argument("--threads", type=optional_int, default=0, help="number of threads used by torch for CPU inference; supercedes MKL_NUM_THREADS/OMP_NUM_THREADS")
+    parser.add_argument("--threads", type=optional_int, default=0, help="number of threads used by torch for CPU inference; supersedes MKL_NUM_THREADS/OMP_NUM_THREADS")
     parser.add_argument("--clip_timestamps", type=str, default="0", help="comma-separated list start,end,start,end,... timestamps (in seconds) of clips to process, where the last end timestamp defaults to the end of the file")
     parser.add_argument("--hallucination_silence_threshold", type=optional_float, help="(requires --word_timestamps True) skip silent periods longer than this threshold (in seconds) when a possible hallucination is detected")
     # fmt: on