np.bool -> bool

hyperion/bin/eval_xvec_cosine_scoring_from_transfer_adv_test_wav.py

@@ -243,7 +243,7 @@ def eval_cosine_scoring(
             vad = v_reader.read([key.seg_set[j]])[0]
             tot_frames = len(vad)
             speech_frames = np.sum(vad)
-            vad = torch.as_tensor(vad.astype(np.bool, copy=False), dtype=torch.bool).to(
+            vad = torch.as_tensor(vad.astype(bool, copy=False), dtype=torch.bool).to(
                 device
             )
             model.vad_t = vad
@@ -361,9 +361,7 @@ def main():
 
     parser.add_argument("--vad", dest="vad_spec", default=None)
     parser.add_argument(
-        "--vad-path-prefix",
-        default=None,
-        help=("scp file_path prefix for vad"),
+        "--vad-path-prefix", default=None, help=("scp file_path prefix for vad"),
     )
 
     parser.add_argument("--model-path", required=True)

hyperion/bin/generate_adv_attacks_xvector_classif.py

@@ -208,18 +208,13 @@ def generate_attacks(
             vad = v_reader.read([key])[0]
             tot_frames = len(vad)
             speech_frames = np.sum(vad)
-            vad = torch.as_tensor(vad.astype(np.bool, copy=False), dtype=torch.bool).to(
+            vad = torch.as_tensor(vad.astype(bool, copy=False), dtype=torch.bool).to(
                 device
             )
             model.vad = vad
             logging.info(
                 "utt %s detected %d/%d (%.2f %%) speech frames"
-                % (
-                    key,
-                    speech_frames,
-                    tot_frames,
-                    speech_frames / tot_frames * 100,
-                )
+                % (key, speech_frames, tot_frames, speech_frames / tot_frames * 100,)
             )
 
         t2 = time.time()
@@ -329,9 +324,7 @@ def main():
 
     parser.add_argument("--vad", dest="vad_spec", default=None)
     parser.add_argument(
-        "--vad-path-prefix",
-        default=None,
-        help=("scp file_path prefix for vad"),
+        "--vad-path-prefix", default=None, help=("scp file_path prefix for vad"),
     )
 
     parser.add_argument("--model-path", required=True)

hyperion/bin/generate_adv_attacks_xvector_verif.py

@@ -217,7 +217,7 @@ def generate_attacks(
             vad = v_reader.read([key.seg_set[j]])[0]
             tot_frames = len(vad)
             speech_frames = np.sum(vad)
-            vad = torch.as_tensor(vad.astype(np.bool, copy=False), dtype=torch.bool).to(
+            vad = torch.as_tensor(vad.astype(bool, copy=False), dtype=torch.bool).to(
                 device
             )
             model.vad_t = vad

hyperion/np/classifiers/greedy_fusion.py

@@ -226,7 +226,7 @@ def fit(self, x, class_ids, sample_weights=None):
             num_cands = len(cand_systems)
             cand_min_dcf = np.zeros((num_cands,), dtype=float_cpu())
             cand_act_dcf = np.zeros((num_cands,), dtype=float_cpu())
-            all_pos = np.zeros((num_cands,), dtype=np.bool)
+            all_pos = np.zeros((num_cands,), dtype=bool)
             cand_weights = []
             for j in range(num_cands):
                 system_idx_ij = np.concatenate(

hyperion/np/diarization/diar_ahc_plda.py

@@ -66,7 +66,7 @@ def _plot_score_hist(scores, output_file, thr=None, gmm=None):
         output_dir = Path(output_file).parent
         output_dir.mkdir(parents=True, exist_ok=True)
 
-        mask = np.triu(np.ones(scores.shape, dtype=np.bool), 1)
+        mask = np.triu(np.ones(scores.shape, dtype=bool), 1)
         scores_r = scores[mask].ravel()
 
         _, bins, _ = plt.hist(
@@ -96,7 +96,7 @@ def _plot_score_hist(scores, output_file, thr=None, gmm=None):
     @staticmethod
     def _unsup_gmm_calibration(scores):
         """Performs unsupervised calibration on the scores by training a GMM."""
-        mask = np.triu(np.ones(scores.shape, dtype=np.bool), 1)
+        mask = np.triu(np.ones(scores.shape, dtype=bool), 1)
         scores_r = scores[mask].ravel()[:, None]  # N x 1
         gmm_1c = GMM(num_comp=1)
         gmm_1c.fit(scores_r, epochs=1)

hyperion/np/pdfs/jfa/jfa_total.py

@@ -281,7 +281,7 @@ def TT(self):
     def _upptr(self):
         """Upper triangular mask."""
         if self.__upptr is None:
-            self.__upptr = np.triu(np.ones(self.y_dim, dtype=np.bool))
+            self.__upptr = np.triu(np.ones(self.y_dim, dtype=bool))
         return self.__upptr
 
     @staticmethod

hyperion/utils/info_table.py

@@ -347,7 +347,7 @@ def get_loc(self, keys):
         if isinstance(loc, int):
             return loc
 
-        if isinstance(loc, np.ndarray) and loc.dtype == np.bool:
+        if isinstance(loc, np.ndarray) and loc.dtype == bool:
             return np.nonzero(loc)[0]
 
         return list(range(loc.start, loc.stop, loc.step))

hyperion/utils/rttm.py

@@ -636,7 +636,7 @@ def get_bin_sample_mask_for_spk(
 
         tend[tend > max_samples] = max_samples
 
-        vad = np.zeros((max_samples,), dtype=np.bool)
+        vad = np.zeros((max_samples,), dtype=bool)
         for i, j in zip(tbeg, tend):
             if j > i:
                 vad[i:j] = True

hyperion/utils/vad_utils.py

@@ -135,7 +135,7 @@ def vad_timestamps_to_bin(
     if max_frames is not None and num_frames < max_frames:
         num_frames = max_frames
 
-    vad = np.zeros((num_frames,), dtype=np.bool)
+    vad = np.zeros((num_frames,), dtype=bool)
     frame_start = np.ceil(
         (in_timestamps[:, 0] - (pad + frame_center)) / frame_shift
     ).astype(dtype=np.int)
@@ -242,7 +242,7 @@ def intersect_segment_timestamps_with_vad(in_timestamps, vad_timestamps):
     vad_start = vad_timestamps[:, 0]
     vad_end = vad_timestamps[:, 1]
     num_vad_segs = len(vad_start)
-    speech_idx = np.zeros((in_timestamps.shape[0],), dtype=np.bool)
+    speech_idx = np.zeros((in_timestamps.shape[0],), dtype=bool)
     out_timestamps = []
     out_timestamps2speech_segs = []
     count_speech = 0