Initial commit

Author: WrathOfGrapes

.gitignore

@@ -0,0 +1,111 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+env/
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# pyenv
+.python-version
+
+# celery beat schedule file
+celerybeat-schedule
+
+# SageMath parsed files
+*.sage.py
+
+# dotenv
+.env
+
+# virtualenv
+.venv
+venv/
+ENV/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+
+# input data, saved log, checkpoints
+data/
+input/
+saved/
+
+# editor, os cache directory
+.vscode/
+.idea/
+__MACOSX/

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2021 Daniil Ivanov
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -0,0 +1,32 @@
+# ASR project barebones
+
+## Installation guide
+
+< Write your installation guide here >
+
+```shell
+pip install -r ./requirements.txt
+```
+
+## Implementation guide
+
+1) Search project for `raise NotImplementedError` and implement missing functionality
+2) Make sure all tests work without errors
+   ```shell
+   python -m unittest discover hw_asr/tests
+   ```
+
+3) Make sure `test.py` works fine and works as expected.
+   You should create files `default_test_config.json` and your installation guide should download your model
+   checpoint and configs in `default_test_model/checkpoint.pth` and `default_test_model/config.json`.
+   ```shell
+   python test.py \
+      -c default_test_config.json \
+      -r default_test_model/checkpoint.pth \
+      -t test_data \
+      -o test_result.json
+   ```
+4) Use `train.py` for training
+
+## Credits
+this repository is based on a heavily modified fork of [pytorch-template](https://github.com/victoresque/pytorch-template) repository.

hw_asr/__init__.py

@@ -0,0 +1,33 @@
+from collections import Callable
+from typing import List
+
+import hw_asr.augmentations.spectrogram_augmentations
+import hw_asr.augmentations.wave_augmentations
+from hw_asr.augmentations.sequential import SequentialAugmentation
+from hw_asr.utils.parse_config import ConfigParser
+
+
+def from_configs(configs: ConfigParser):
+    wave_augs = []
+    if "augmentations" in configs.config and "wave" in configs.config["augmentations"]:
+        for aug_dict in configs.config["augmentations"]["wave"]:
+            wave_augs.append(
+                configs.init_obj(aug_dict, hw_asr.augmentations.wave_augmentations)
+            )
+
+    spec_augs = []
+    if "augmentations" in configs.config and "spectrogram" in configs.config["augmentations"]:
+        for aug_dict in configs.config["augmentations"]["spectrogram"]:
+            spec_augs.append(
+                configs.init_obj(aug_dict, hw_asr.augmentations.spectrogram_augmentations)
+            )
+    return _to_function(wave_augs), _to_function(spec_augs)
+
+
+def _to_function(augs_list: List[Callable]):
+    if len(augs_list) == 0:
+        return None
+    elif len(augs_list) == 1:
+        return augs_list[0]
+    else:
+        return SequentialAugmentation(augs_list)

hw_asr/augmentations/__init__.py

@@ -0,0 +1,6 @@
+from torch import Tensor
+
+
+class AugmentationBase:
+    def __call__(self, data: Tensor) -> Tensor:
+        raise NotImplementedError

hw_asr/augmentations/base.py

@@ -0,0 +1,17 @@
+import random
+from typing import Callable
+
+from torch import Tensor
+
+
+class RandomApply:
+    def __init__(self, augmentation: Callable, p: float):
+        assert 0 <= p <= 1
+        self.augmentation = augmentation
+        self.p = p
+
+    def __call__(self, data: Tensor) -> Tensor:
+        if random.random() < self.p:
+            return self.augmentation(data)
+        else:
+            return data

hw_asr/augmentations/random_apply.py

@@ -0,0 +1,16 @@
+from typing import List, Callable
+
+from torch import Tensor
+
+from hw_asr.augmentations.base import AugmentationBase
+
+
+class SequentialAugmentation(AugmentationBase):
+    def __init__(self, augmentation_list: List[Callable]):
+        self.augmentation_list = augmentation_list
+
+    def __call__(self, data: Tensor) -> Tensor:
+        x = data
+        for augmentation in self.augmentation_list:
+            x = augmentation(data)
+        return x

hw_asr/augmentations/sequential.py

@@ -0,0 +1 @@
+

hw_asr/augmentations/spectrogram_augmentations/__init__.py

@@ -0,0 +1,13 @@
+import torch_audiomentations
+from torch import Tensor
+
+from hw_asr.augmentations.base import AugmentationBase
+
+
+class Gain(AugmentationBase):
+    def __init__(self, *args, **kwargs):
+        self._aug = torch_audiomentations.Gain(*args, **kwargs)
+
+    def __call__(self, data: Tensor):
+        x = data.unsqueeze(1)
+        return self._aug(x).squeeze(1)

hw_asr/augmentations/wave_augmentations/Gain.py

@@ -0,0 +1,5 @@
+from hw_asr.augmentations.wave_augmentations.Gain import Gain
+
+__all__ = [
+    "Gain"
+]

hw_asr/augmentations/wave_augmentations/__init__.py

@@ -0,0 +1,2 @@
+from .base_model import *
+from .base_trainer import *