Update to latest linters

.flake8

@@ -1,7 +1,7 @@
 [flake8]
-ignore = E203, E266, E501, W503
+ignore = E203, E266, E501, W503, DOC105, DOC106, DOC107, DOC203, DOC403, DOC404, DOC405, DOC501, DOC502
 # line length is intentionally set to 80 here because black uses Bugbear
 # See https://github.com/psf/black/blob/master/README.md#line-length for more details
 max-line-length = 80
 max-complexity = 18
-select = B,C,E,F,W,T4,B9
+select = B,C,E,F,W,T4,B9,DOC

.pre-commit-config.yaml

@@ -14,18 +14,26 @@ repos:
         args: [--maxkb=1024]
 
   - repo: https://github.com/psf/black
-    rev: 19.10b0
+    rev: 24.3.0
     hooks:
       - id: black
         types: [python]
-        additional_dependencies: ['click==8.0.4']
+        additional_dependencies: ['click==8.1.7']
   - repo: https://github.com/PyCQA/flake8
-    rev: 3.7.9
+    rev: 7.0.0
     hooks:
       - id: flake8
         types: [python]
 
   - repo: https://github.com/adrienverge/yamllint
-    rev: v1.23.0
+    rev: v1.35.1
     hooks:
       - id: yamllint
+
+  - repo: https://github.com/codespell-project/codespell
+    rev: v2.2.4
+    hooks:
+      - id: codespell
+        args: [--ignore-words=.dict-speechbrain.txt]
+        additional_dependencies:
+          - tomli

benchmarks/CL_MASR/analyze_logs.py

@@ -64,13 +64,13 @@ def parse_train_log(train_log: "str") -> "Dict[str, ndarray]":
 
     Arguments
     ---------
-    train_log:
+    train_log: str
         The path to the train log.
 
     Returns
     -------
-        The metrics, i.e. a dict that maps names of
-        the metrics to their corresponding values.
+    The metrics, i.e. a dict that maps names of
+    the metrics to their corresponding values.
 
     Examples
     --------
@@ -107,16 +107,16 @@ def compute_wer_matrix(
 
     Arguments
     ---------
-    wers:
+    wers: ndarray
         The word error rate for each locale.
-    num_base_locales:
+    num_base_locales: int
         The number of base locales.
-    num_new_locales:
+    num_new_locales: int
         The number of new locales.
 
     Returns
     -------
-        The word error rate matrix.
+    The word error rate matrix.
 
     Raises
     ------
@@ -152,12 +152,12 @@ def compute_awer(wer_matrix: "ndarray") -> "ndarray":
 
     Arguments
     ---------
-    wer_matrix:
+    wer_matrix: ndarray
         The word error rate matrix.
 
     Returns
     -------
-        The average word error rate.
+    The average word error rate.
 
     References
     ----------
@@ -185,12 +185,12 @@ def compute_bwt(wer_matrix: "ndarray") -> "ndarray":
 
     Arguments
     ---------
-    wer_matrix:
+    wer_matrix: ndarray
         The word error rate matrix.
 
     Returns
     -------
-        The backward transfer.
+    The backward transfer.
 
     References
     ----------
@@ -220,14 +220,14 @@ def compute_im(wer_matrix: "ndarray", refs: "ndarray") -> "ndarray":
 
     Arguments
     ---------
-    wer_matrix:
+    wer_matrix: ndarray
         The word error rate matrix.
-    refs:
+    refs: ndarray
         The intransigence measure references (joint fine-tuning).
 
     Returns
     -------
-        The intransigence measure.
+    The intransigence measure.
 
     References
     ----------
@@ -255,14 +255,14 @@ def compute_fwt(wer_matrix: "ndarray", refs: "ndarray") -> "ndarray":
 
     Arguments
     ---------
-    wer_matrix:
+    wer_matrix: ndarray
         The word error rate matrix.
-    refs:
+    refs: ndarray
         The forward transfer references (single task fine-tuning).
 
     Returns
     -------
-        The forward transfer.
+    The forward transfer.
 
     Examples
     --------
@@ -289,31 +289,31 @@ def plot_wer(
     usetex: "bool" = False,
     hide_legend: "bool" = False,
     style_file_or_name: "str" = "classic",
-) -> "None":
+):
     """Plot word error rates extracted from a
     continual learning train log.
 
     Arguments
     ---------
-    wers:
+    wers: ndarray
         The word error rates (base + new locales).
-    output_image:
+    output_image: str
         The path to the output image.
-    base_locales:
+    base_locales: Sequence[str]
         The base locales.
-    new_locales:
+    new_locales: Sequence[str]
         The new locales.
-    xlabel:
+    xlabel: str
         The x-axis label.
-    figsize:
+    figsize: Tuple[float, float]
         The figure size.
-    title:
+    title: str
         The plot title.
-    usetex:
+    usetex: bool
         True to render text with LaTeX, False otherwise.
-    hide_legend:
+    hide_legend: bool
         True to hide the legend, False otherwise.
-    style_file_or_name:
+    style_file_or_name: str
         The path to a Matplotlib style file or the name of one
         of Matplotlib built-in styles
         (see https://matplotlib.org/stable/gallery/style_sheets/style_sheets_reference.html).
@@ -435,7 +435,7 @@ def plot_wer(
             margin={"t": 60, "b": 60},
         )
         fig.write_html(
-            f"{output_image.rsplit('.', 1)[0]}.html", include_plotlyjs=True,
+            f"{output_image.rsplit('.', 1)[0]}.html", include_plotlyjs=True
         )
     except ImportError:
         logging.warning(
@@ -455,32 +455,32 @@ def plot_metric(
     usetex: "bool" = False,
     hide_legend: "bool" = False,
     style_file_or_name: "str" = "classic",
-) -> "None":
+):
     """Plot a continual learning metric.
 
     Arguments
     ---------
-    metric_csv_file:
+    metric_csv_file: str
         The path to the continual learning metric CSV file.
-    output_image:
+    output_image: str
         The path to the output image.
-    xlabel:
+    xlabel: str
         The x-axis label.
-    ylabel:
+    ylabel: str
         The y-axis label.
-    xticks:
+    xticks: List[str]
         The x-ticks.
-    figsize:
+    figsize: Tuple[float, float]
         The figure size.
-    title:
+    title: str
         The plot title.
-    opacity:
+    opacity: float
         The confidence interval opacity.
-    usetex:
+    usetex: bool
         True to render text with LaTeX, False otherwise.
-    hide_legend:
+    hide_legend: bool
         True to hide the legend, False otherwise.
-    style_file_or_name:
+    style_file_or_name: str
         The path to a Matplotlib style file or the name of one
         of Matplotlib built-in styles
         (see https://matplotlib.org/stable/gallery/style_sheets/style_sheets_reference.html).
@@ -650,7 +650,7 @@ def hex_to_rgb(hex_color: "str") -> "Tuple":
             margin={"t": 60, "b": 60},
         )
         fig.write_html(
-            f"{output_image.rsplit('.', 1)[0]}.html", include_plotlyjs=True,
+            f"{output_image.rsplit('.', 1)[0]}.html", include_plotlyjs=True
         )
     except ImportError:
         logging.warning(
@@ -689,9 +689,7 @@ def hex_to_rgb(hex_color: "str") -> "Tuple":
         # fmt: on
         help="forward transfer references",
     )
-    parser.add_argument(
-        "-f", "--format", default="png", help="image format",
-    )
+    parser.add_argument("-f", "--format", default="png", help="image format")
     parser.add_argument(
         "-s",
         "--figsize",
@@ -700,17 +698,15 @@ def hex_to_rgb(hex_color: "str") -> "Tuple":
         type=float,
         help="figure size",
     )
+    parser.add_argument("-t", "--title", default=None, help="title")
     parser.add_argument(
-        "-t", "--title", default=None, help="title",
-    )
-    parser.add_argument(
-        "-o", "--opacity", default=0.15, help="confidence interval opacity",
+        "-o", "--opacity", default=0.15, help="confidence interval opacity"
     )
     parser.add_argument(
-        "--hide_legend", action="store_true", help="hide legend",
+        "--hide_legend", action="store_true", help="hide legend"
     )
     parser.add_argument(
-        "-u", "--usetex", action="store_true", help="render text with LaTeX",
+        "-u", "--usetex", action="store_true", help="render text with LaTeX"
     )
     parser.add_argument(
         "--order",
@@ -831,7 +827,7 @@ def hex_to_rgb(hex_color: "str") -> "Tuple":
                 avg_mean = np.mean(avg)
                 # Assuming independence, sigma^2 = sum_1^n sigma_i^2 / n^2
                 avg_stddev = np.sqrt(
-                    np.nansum(stddev ** 2) / (~np.isnan(stddev)).sum() ** 2
+                    np.nansum(stddev**2) / (~np.isnan(stddev)).sum() ** 2
                 )
                 csv_writer.writerow(
                     [group_name]
@@ -851,9 +847,9 @@ def hex_to_rgb(hex_color: "str") -> "Tuple":
                 f"{name.lower().replace(' ', '_')}.{args.format}",
             ),
             xlabel=None,
-            ylabel=f"{name} (\%)"
-            if args.usetex
-            else f"{name} (%)",  # noqa: W605
+            ylabel=(
+                f"{name} (\\%)" if args.usetex else f"{name} (%)"
+            ),  # noqa: W605
             xticks=["base"] + [f"L{i}" for i in range(1, 1 + len(new_locales))],
             figsize=args.figsize,
             title=args.title,