sonusai 1.0.2__tar.gz → 1.0.4__tar.gz

{sonusai-1.0.2 → sonusai-1.0.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: sonusai
-Version: 1.0.2
+Version: 1.0.4
 Summary: Framework for building deep neural network models for sound, speech, and voice AI
 Home-page: https://aaware.com
 License: GPL-3.0-only
@@ -36,6 +36,7 @@ Requires-Dist: pyyaml (>=6.0.2,<7.0.0)
 Requires-Dist: requests (>=2.32.3,<3.0.0)
 Requires-Dist: rich (>=13.9.4,<14.0.0)
 Requires-Dist: samplerate (>=0.2.1,<0.3.0)
+Requires-Dist: scikit-learn (>=1.6.1,<2.0.0)
 Requires-Dist: sh (>=2.2.2,<3.0.0)
 Requires-Dist: soundfile (>=0.12.1,<0.13.0)
 Requires-Dist: sox (>=1.5.0,<2.0.0)

{sonusai-1.0.2 → sonusai-1.0.4}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "sonusai"
-version = "1.0.2"
+version = "1.0.4"
 description = "Framework for building deep neural network models for sound, speech, and voice AI"
 authors = ["Chris Eddington <chris@aaware.com>", "Jason Calderwood <jason@aaware.com>"]
 maintainers = ["Chris Eddington <chris@aaware.com>", "Jason Calderwood <jason@aaware.com>"]
@@ -42,6 +42,7 @@ pyyaml = "^6.0.2"
 requests = "^2.32.3"
 rich = "^13.9.4"
 samplerate = "^0.2.1"
+scikit-learn = "^1.6.1"
 sh = "^2.2.2"
 soundfile = "^0.12.1"
 sox = "^1.5.0"

{sonusai-1.0.2 → sonusai-1.0.4}/sonusai/genmixdb.py

@@ -91,6 +91,9 @@ def genmixdb(
     if len([file for file in source_files if file.category == "primary"]) == 0:
         raise RuntimeError("Canceled due to no primary sources")
 
+    if logging:
+        logger.info("Populating source file table")
+
     populate_source_file_table(location, source_files, test)
 
     if logging:
@@ -123,6 +126,9 @@ def genmixdb(
     ir_files = get_ir_files(config, show_progress=show_progress)
     logger.info("")
 
+    if logging:
+        logger.info("Populating impulse response file table")
+
     populate_impulse_response_file_table(location, ir_files, test)
 
     if logging:

{sonusai-1.0.2 → sonusai-1.0.4}/sonusai/mixture/effects.py

@@ -308,7 +308,7 @@ def evaluate_sai_random_ir(mixdb: MixtureDatabase, text: str) -> str:
         return m.group(1)
 
     if re.match(rand_pattern, text):
-        return f"ir {randint(0, mixdb.num_ir_files)}"  # noqa: S311
+        return f"ir {randint(0, mixdb.num_ir_files - 1)}"  # noqa: S311
 
     if re.match(rand_range_pattern, text):
         try:

{sonusai-1.0.2 → sonusai-1.0.4}/sonusai/mixture/generation.py

@@ -437,6 +437,7 @@ def populate_mixture_table(
     # Populate source table
     if logging:
         logger.info("Populating source table")
+    # TODO: refactor this to not load all sources into list; maybe us UNIQUE table modifier?
     sources: list[tuple[str, int, float, bool, float, float, bool, int]] = []
     for mixture in mixtures:
         for source in mixture.all_sources.values():
@@ -577,7 +578,7 @@ def _initialize_mixture_gains(mixdb: MixtureDatabase, mixture: Mixture, sources_
             # Special case for zeroing out noise data
             mixture.all_sources[category].snr_gain = 0
         elif category != "primary":
-            if sources_energy["primary"] == 0:
+            if sources_energy["primary"] == 0 or sources_energy[category] == 0:
                 # Avoid divide-by-zero
                 mixture.all_sources[category].snr_gain = 1
             else: