Spaces:
Runtime error
Runtime error
work with grayscale images
Browse files- notebooks/test-model.ipynb +0 -0
- requirements-lock.txt +198 -0
- src/audio_to_images.py +5 -0
- src/train_unconditional.py +4 -5
notebooks/test-model.ipynb
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
requirements-lock.txt
ADDED
@@ -0,0 +1,198 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
absl-py==1.2.0
|
2 |
+
accelerate==0.12.0
|
3 |
+
aiobotocore==2.3.4
|
4 |
+
aiohttp==3.8.1
|
5 |
+
aioitertools==0.10.0
|
6 |
+
aiosignal==1.2.0
|
7 |
+
analytics-python==1.4.0
|
8 |
+
anyio==3.6.1
|
9 |
+
appdirs==1.4.4
|
10 |
+
argon2-cffi==21.3.0
|
11 |
+
argon2-cffi-bindings==21.2.0
|
12 |
+
async-timeout==4.0.2
|
13 |
+
attrs==21.4.0
|
14 |
+
audioread==2.1.9
|
15 |
+
backcall==0.2.0
|
16 |
+
backoff==1.10.0
|
17 |
+
bcrypt==3.2.2
|
18 |
+
beautifulsoup4==4.11.1
|
19 |
+
bertviz==1.4.0
|
20 |
+
black==22.6.0
|
21 |
+
bleach==5.0.1
|
22 |
+
boto3==1.21.21
|
23 |
+
botocore==1.24.21
|
24 |
+
cachetools==5.2.0
|
25 |
+
captum==0.5.0
|
26 |
+
certifi==2022.6.15
|
27 |
+
cffi==1.15.1
|
28 |
+
charset-normalizer==2.1.0
|
29 |
+
click==8.1.3
|
30 |
+
cloudpickle==2.1.0
|
31 |
+
cryptography==37.0.4
|
32 |
+
cycler==0.11.0
|
33 |
+
datasets==2.4.0
|
34 |
+
debugpy==1.6.2
|
35 |
+
decorator==5.1.1
|
36 |
+
deepspeed==0.7.0
|
37 |
+
defusedxml==0.7.1
|
38 |
+
diffusers==0.1.3
|
39 |
+
dill==0.3.5.1
|
40 |
+
entrypoints==0.4
|
41 |
+
fastapi==0.79.0
|
42 |
+
fastjsonschema==2.16.1
|
43 |
+
ffmpy==0.3.0
|
44 |
+
filelock==3.8.0
|
45 |
+
fonttools==4.34.4
|
46 |
+
frozenlist==1.3.1
|
47 |
+
fsspec==2022.7.1
|
48 |
+
google-auth==2.10.0
|
49 |
+
google-auth-oauthlib==0.4.6
|
50 |
+
google-pasta==0.2.0
|
51 |
+
gradio==3.1.4
|
52 |
+
grpcio==1.47.0
|
53 |
+
h11==0.12.0
|
54 |
+
hjson==3.0.2
|
55 |
+
httpcore==0.15.0
|
56 |
+
httpx==0.23.0
|
57 |
+
huggingface-hub==0.8.1
|
58 |
+
idna==3.3
|
59 |
+
importlib-metadata==4.12.0
|
60 |
+
ipykernel==6.15.1
|
61 |
+
ipython==7.34.0
|
62 |
+
ipython-genutils==0.2.0
|
63 |
+
ipywidgets==7.7.1
|
64 |
+
jedi==0.18.1
|
65 |
+
Jinja2==3.1.2
|
66 |
+
jmespath==1.0.1
|
67 |
+
joblib==1.1.0
|
68 |
+
jsonschema==4.9.1
|
69 |
+
jupyter-client==7.3.4
|
70 |
+
jupyter-core==4.11.1
|
71 |
+
jupyterlab-pygments==0.2.2
|
72 |
+
jupyterlab-widgets==1.1.1
|
73 |
+
kiwisolver==1.4.4
|
74 |
+
librosa==0.9.2
|
75 |
+
linkify-it-py==1.0.3
|
76 |
+
llvmlite==0.39.0
|
77 |
+
lxml==4.9.1
|
78 |
+
Markdown==3.4.1
|
79 |
+
markdown-it-py==2.1.0
|
80 |
+
MarkupSafe==2.1.1
|
81 |
+
matplotlib==3.5.2
|
82 |
+
matplotlib-inline==0.1.3
|
83 |
+
mdit-py-plugins==0.3.0
|
84 |
+
mdurl==0.1.1
|
85 |
+
mistune==0.8.4
|
86 |
+
monotonic==1.6
|
87 |
+
more-itertools==8.14.0
|
88 |
+
multidict==6.0.2
|
89 |
+
multiprocess==0.70.13
|
90 |
+
munkres==1.1.4
|
91 |
+
mypy-extensions==0.4.3
|
92 |
+
nbclient==0.6.6
|
93 |
+
nbconvert==6.5.1
|
94 |
+
nbformat==5.4.0
|
95 |
+
nest-asyncio==1.5.5
|
96 |
+
networkx==2.8.5
|
97 |
+
ninja==1.10.2.3
|
98 |
+
nlp==0.4.0
|
99 |
+
nltk==3.7
|
100 |
+
notebook==6.4.12
|
101 |
+
numba==0.56.0
|
102 |
+
numpy==1.22.4
|
103 |
+
oauthlib==3.2.0
|
104 |
+
orjson==3.7.11
|
105 |
+
packaging==21.3
|
106 |
+
pandas==1.4.3
|
107 |
+
pandocfilters==1.5.0
|
108 |
+
paramiko==2.11.0
|
109 |
+
parso==0.8.3
|
110 |
+
pathos==0.2.9
|
111 |
+
pathspec==0.9.0
|
112 |
+
pexpect==4.8.0
|
113 |
+
pickleshare==0.7.5
|
114 |
+
Pillow==9.2.0
|
115 |
+
platformdirs==2.5.2
|
116 |
+
pluggy==0.13.1
|
117 |
+
pooch==1.6.0
|
118 |
+
pox==0.3.1
|
119 |
+
ppft==1.7.6.5
|
120 |
+
prometheus-client==0.14.1
|
121 |
+
prompt-toolkit==3.0.30
|
122 |
+
protobuf==3.19.4
|
123 |
+
protobuf3-to-dict==0.1.5
|
124 |
+
psutil==5.9.1
|
125 |
+
ptyprocess==0.7.0
|
126 |
+
py==1.11.0
|
127 |
+
py-cpuinfo==8.0.0
|
128 |
+
pyarrow==9.0.0
|
129 |
+
pyasn1==0.4.8
|
130 |
+
pyasn1-modules==0.2.8
|
131 |
+
pycparser==2.21
|
132 |
+
pycryptodome==3.15.0
|
133 |
+
pydantic==1.9.1
|
134 |
+
pydub==0.25.1
|
135 |
+
Pygments==2.12.0
|
136 |
+
PyNaCl==1.5.0
|
137 |
+
pyparsing==3.0.9
|
138 |
+
pyrsistent==0.18.1
|
139 |
+
pytest==5.4.3
|
140 |
+
python-dateutil==2.8.2
|
141 |
+
python-dotenv==0.20.0
|
142 |
+
python-multipart==0.0.5
|
143 |
+
pytz==2022.1
|
144 |
+
PyYAML==6.0
|
145 |
+
pyzmq==23.2.0
|
146 |
+
regex==2022.7.25
|
147 |
+
requests==2.28.1
|
148 |
+
requests-oauthlib==1.3.1
|
149 |
+
resampy==0.4.0
|
150 |
+
responses==0.18.0
|
151 |
+
rfc3986==1.5.0
|
152 |
+
rsa==4.9
|
153 |
+
s3fs==2022.7.1
|
154 |
+
s3transfer==0.5.2
|
155 |
+
sagemaker==2.103.0
|
156 |
+
scikit-learn==1.1.2
|
157 |
+
scipy==1.9.0
|
158 |
+
seaborn==0.11.2
|
159 |
+
Send2Trash==1.8.0
|
160 |
+
sentencepiece==0.1.97
|
161 |
+
shap==0.41.0
|
162 |
+
six==1.16.0
|
163 |
+
slicer==0.0.7
|
164 |
+
smdebug-rulesconfig==1.0.1
|
165 |
+
sniffio==1.2.0
|
166 |
+
snorkel==0.9.9
|
167 |
+
SoundFile==0.10.3.post1
|
168 |
+
soupsieve==2.3.2.post1
|
169 |
+
starlette==0.19.1
|
170 |
+
tensorboard==2.9.1
|
171 |
+
tensorboard-data-server==0.6.1
|
172 |
+
tensorboard-plugin-wit==1.8.1
|
173 |
+
terminado==0.15.0
|
174 |
+
threadpoolctl==3.1.0
|
175 |
+
tinycss2==1.1.1
|
176 |
+
tokenizers==0.12.1
|
177 |
+
toml==0.10.2
|
178 |
+
tomli==2.0.1
|
179 |
+
torch==1.12.1
|
180 |
+
torchvision==0.13.1
|
181 |
+
tornado==6.2
|
182 |
+
tqdm==4.64.0
|
183 |
+
traitlets==5.3.0
|
184 |
+
transformers==4.21.1
|
185 |
+
transformers-interpret==0.7.5
|
186 |
+
typing_extensions==4.3.0
|
187 |
+
uc-micro-py==1.0.1
|
188 |
+
urllib3==1.26.11
|
189 |
+
uvicorn==0.18.2
|
190 |
+
wcwidth==0.2.5
|
191 |
+
webencodings==0.5.1
|
192 |
+
Werkzeug==2.2.2
|
193 |
+
widgetsnbextension==3.6.1
|
194 |
+
wrapt==1.14.1
|
195 |
+
xxhash==3.0.0
|
196 |
+
yapf==0.32.0
|
197 |
+
yarl==1.8.1
|
198 |
+
zipp==3.8.1
|
src/audio_to_images.py
CHANGED
@@ -1,3 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import os
|
2 |
import re
|
3 |
import io
|
|
|
1 |
+
# TODO
|
2 |
+
# run on sagemaker
|
3 |
+
# run with deepspeed
|
4 |
+
|
5 |
+
|
6 |
import os
|
7 |
import re
|
8 |
import io
|
src/train_unconditional.py
CHANGED
@@ -39,8 +39,8 @@ def main(args):
|
|
39 |
|
40 |
model = UNet2DModel(
|
41 |
sample_size=args.resolution,
|
42 |
-
in_channels=
|
43 |
-
out_channels=
|
44 |
layers_per_block=2,
|
45 |
block_out_channels=(128, 128, 256, 256, 512, 512),
|
46 |
down_block_types=(
|
@@ -101,7 +101,7 @@ def main(args):
|
|
101 |
)
|
102 |
|
103 |
def transforms(examples):
|
104 |
-
images = [augmentations(image
|
105 |
return {"input": images}
|
106 |
|
107 |
dataset.set_transform(transforms)
|
@@ -215,8 +215,7 @@ def main(args):
|
|
215 |
"test_samples", images_processed, epoch
|
216 |
)
|
217 |
for _, image in enumerate(images_processed):
|
218 |
-
|
219 |
-
audio = mel.image_to_audio(image)
|
220 |
accelerator.trackers[0].writer.add_audio(
|
221 |
f"test_audio_{_}",
|
222 |
audio,
|
|
|
39 |
|
40 |
model = UNet2DModel(
|
41 |
sample_size=args.resolution,
|
42 |
+
in_channels=1,
|
43 |
+
out_channels=1,
|
44 |
layers_per_block=2,
|
45 |
block_out_channels=(128, 128, 256, 256, 512, 512),
|
46 |
down_block_types=(
|
|
|
101 |
)
|
102 |
|
103 |
def transforms(examples):
|
104 |
+
images = [augmentations(image) for image in examples["image"]]
|
105 |
return {"input": images}
|
106 |
|
107 |
dataset.set_transform(transforms)
|
|
|
215 |
"test_samples", images_processed, epoch
|
216 |
)
|
217 |
for _, image in enumerate(images_processed):
|
218 |
+
audio = mel.image_to_audio(Image.fromarray(image[0]))
|
|
|
219 |
accelerator.trackers[0].writer.add_audio(
|
220 |
f"test_audio_{_}",
|
221 |
audio,
|