Skip to content

Commit ec263cb

Browse files
author
Github Actions
committed
nabenabe0928: [feat] Add flexible step-wise LR scheduler with minimum changes (#256)
1 parent 256c4bc commit ec263cb

24 files changed

+666
-376
lines changed

‎development/_downloads/bc82bea3a5dd7bdba60b65220891d9e5/examples_python.zip‎

0 Bytes
Binary file not shown.

‎development/_downloads/fb625db3c50d423b1b7881136ffdeec8/examples_jupyter.zip‎

0 Bytes
Binary file not shown.

‎development/_images/sphx_glr_example_visualization_001.png‎

-1.87 KB
Loading

‎development/_images/sphx_glr_example_visualization_002.png‎

3 Bytes
Loading

‎development/_images/sphx_glr_example_visualization_thumb.png‎

98 Bytes
Loading

‎development/_sources/examples/20_basics/example_image_classification.rst.txt‎

Lines changed: 6 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -86,17 +86,12 @@ Image Classification
8686
________________________________________
8787
Configuration:
8888
image_augmenter:GaussianBlur:use_augmenter, Value: False
89-
image_augmenter:GaussianNoise:sigma_offset, Value: 2.103193881188938
90-
image_augmenter:GaussianNoise:use_augmenter, Value: True
91-
image_augmenter:RandomAffine:rotate, Value: 128
92-
image_augmenter:RandomAffine:scale_offset, Value: 0.04320130291561389
93-
image_augmenter:RandomAffine:shear, Value: 28
94-
image_augmenter:RandomAffine:translate_percent_offset, Value: 0.3444543885349758
95-
image_augmenter:RandomAffine:use_augmenter, Value: True
96-
image_augmenter:RandomCutout:p, Value: 0.2059781936341846
89+
image_augmenter:GaussianNoise:use_augmenter, Value: False
90+
image_augmenter:RandomAffine:use_augmenter, Value: False
91+
image_augmenter:RandomCutout:p, Value: 0.44825795481578556
9792
image_augmenter:RandomCutout:use_augmenter, Value: True
98-
image_augmenter:Resize:use_augmenter, Value: True
99-
image_augmenter:ZeroPadAndCrop:percent, Value: 0.1312339317525968
93+
image_augmenter:Resize:use_augmenter, Value: False
94+
image_augmenter:ZeroPadAndCrop:percent, Value: 0.2662934360012483
10095
normalizer:__choice__, Value: 'ImageNormalizer'
10196

10297
Fitting the pipeline...
@@ -176,7 +171,7 @@ Image Classification
176171
177172
.. rst-class:: sphx-glr-timing
178173

179-
**Total running time of the script:** ( 0 minutes 5.789 seconds)
174+
**Total running time of the script:** ( 0 minutes 6.500 seconds)
180175

181176

182177
.. _sphx_glr_download_examples_20_basics_example_image_classification.py:

‎development/_sources/examples/20_basics/example_tabular_classification.rst.txt‎

Lines changed: 13 additions & 16 deletions
Original file line numberDiff line numberDiff line change
@@ -133,7 +133,7 @@ Search for an ensemble of machine learning algorithms
133133
.. code-block:: none
134134
135135
136-
<autoPyTorch.api.tabular_classification.TabularClassificationTask object at 0x7f23f4a97fa0>
136+
<autoPyTorch.api.tabular_classification.TabularClassificationTask object at 0x7f7f1c3e3340>
137137
138138
139139
@@ -162,7 +162,7 @@ Print the final ensemble performance
162162

163163
.. code-block:: none
164164
165-
<smac.runhistory.runhistory.RunHistory object at 0x7f23f4a97a30> [TrajEntry(train_perf=2147483648, incumbent_id=1, incumbent=Configuration:
165+
<smac.runhistory.runhistory.RunHistory object at 0x7f7f1c3e3d30> [TrajEntry(train_perf=2147483648, incumbent_id=1, incumbent=Configuration:
166166
data_loader:batch_size, Value: 64
167167
encoder:__choice__, Value: 'OneHotEncoder'
168168
feature_preprocessor:__choice__, Value: 'NoFeaturePreprocessor'
@@ -194,7 +194,7 @@ Print the final ensemble performance
194194
scaler:__choice__, Value: 'StandardScaler'
195195
trainer:StandardTrainer:weighted_loss, Value: True
196196
trainer:__choice__, Value: 'StandardTrainer'
197-
, ta_runs=0, ta_time_used=0.0, wallclock_time=0.0012044906616210938, budget=0), TrajEntry(train_perf=0.16374269005847952, incumbent_id=1, incumbent=Configuration:
197+
, ta_runs=0, ta_time_used=0.0, wallclock_time=0.0012297630310058594, budget=0), TrajEntry(train_perf=0.16374269005847952, incumbent_id=1, incumbent=Configuration:
198198
data_loader:batch_size, Value: 64
199199
encoder:__choice__, Value: 'OneHotEncoder'
200200
feature_preprocessor:__choice__, Value: 'NoFeaturePreprocessor'
@@ -226,29 +226,26 @@ Print the final ensemble performance
226226
scaler:__choice__, Value: 'StandardScaler'
227227
trainer:StandardTrainer:weighted_loss, Value: True
228228
trainer:__choice__, Value: 'StandardTrainer'
229-
, ta_runs=1, ta_time_used=1.9000828266143799, wallclock_time=3.1139419078826904, budget=5.555555555555555)]
230-
{'accuracy': 0.8323699421965318}
229+
, ta_runs=1, ta_time_used=1.8366239070892334, wallclock_time=3.0440237522125244, budget=5.555555555555555)]
230+
{'accuracy': 0.861271676300578}
231231
| | Preprocessing | Estimator | Weight |
232232
|---:|:------------------------------------------------------------------|:-------------------------------------------------------------------|---------:|
233-
| 0 | SimpleImputer,NoEncoder,NoScaler,KernelPCA | no embedding,ResNetBackbone,FullyConnectedHead,nn.Sequential | 0.26 |
234-
| 1 | None | RFLearner | 0.12 |
233+
| 0 | SimpleImputer,NoEncoder,MinMaxScaler,Nystroem | no embedding,ShapedMLPBackbone,FullyConnectedHead,nn.Sequential | 0.52 |
234+
| 1 | SimpleImputer,NoEncoder,NoScaler,KernelPCA | no embedding,ResNetBackbone,FullyConnectedHead,nn.Sequential | 0.14 |
235235
| 2 | SimpleImputer,OneHotEncoder,MinMaxScaler,KernelPCA | no embedding,ShapedResNetBackbone,FullyConnectedHead,nn.Sequential | 0.1 |
236-
| 3 | SimpleImputer,OneHotEncoder,MinMaxScaler,PolynomialFeatures | no embedding,ShapedResNetBackbone,FullyConnectedHead,nn.Sequential | 0.1 |
237-
| 4 | None | CBLearner | 0.1 |
238-
| 5 | SimpleImputer,OneHotEncoder,MinMaxScaler,KernelPCA | no embedding,ShapedResNetBackbone,FullyConnectedHead,nn.Sequential | 0.08 |
239-
| 6 | SimpleImputer,OneHotEncoder,NoScaler,TruncSVD | embedding,ShapedResNetBackbone,FullyConnectedHead,nn.Sequential | 0.08 |
240-
| 7 | SimpleImputer,OneHotEncoder,NoScaler,PowerTransformer | no embedding,ShapedMLPBackbone,FullyConnectedHead,nn.Sequential | 0.04 |
241-
| 8 | None | LGBMLearner | 0.04 |
242-
| 9 | None | ETLearner | 0.04 |
243-
| 10 | SimpleImputer,OneHotEncoder,StandardScaler,NoFeaturePreprocessing | no embedding,ShapedMLPBackbone,FullyConnectedHead,nn.Sequential | 0.04 |
236+
| 3 | None | CBLearner | 0.08 |
237+
| 4 | None | RFLearner | 0.08 |
238+
| 5 | SimpleImputer,OneHotEncoder,MinMaxScaler,KernelPCA | no embedding,ShapedResNetBackbone,FullyConnectedHead,nn.Sequential | 0.04 |
239+
| 6 | SimpleImputer,OneHotEncoder,NoScaler,TruncSVD | embedding,ShapedResNetBackbone,FullyConnectedHead,nn.Sequential | 0.02 |
240+
| 7 | SimpleImputer,OneHotEncoder,StandardScaler,NoFeaturePreprocessing | no embedding,ShapedMLPBackbone,FullyConnectedHead,nn.Sequential | 0.02 |
244241
245242
246243
247244
248245
249246
.. rst-class:: sphx-glr-timing
250247

251-
**Total running time of the script:** ( 5 minutes 24.604 seconds)
248+
**Total running time of the script:** ( 5 minutes 25.351 seconds)
252249

253250

254251
.. _sphx_glr_download_examples_20_basics_example_tabular_classification.py:

‎development/_sources/examples/20_basics/example_tabular_regression.rst.txt‎

Lines changed: 7 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -125,7 +125,7 @@ Search for an ensemble of machine learning algorithms
125125
.. code-block:: none
126126
127127
128-
<autoPyTorch.api.tabular_regression.TabularRegressionTask object at 0x7f24a46e6880>
128+
<autoPyTorch.api.tabular_regression.TabularRegressionTask object at 0x7f7fbca046a0>
129129
130130
131131
@@ -157,7 +157,7 @@ Print the final ensemble performance
157157

158158
.. code-block:: none
159159
160-
<smac.runhistory.runhistory.RunHistory object at 0x7f249bb9e3d0> [TrajEntry(train_perf=2147483648, incumbent_id=1, incumbent=Configuration:
160+
<smac.runhistory.runhistory.RunHistory object at 0x7f7fa79e30a0> [TrajEntry(train_perf=2147483648, incumbent_id=1, incumbent=Configuration:
161161
data_loader:batch_size, Value: 64
162162
encoder:__choice__, Value: 'OneHotEncoder'
163163
feature_preprocessor:__choice__, Value: 'NoFeaturePreprocessor'
@@ -188,7 +188,7 @@ Print the final ensemble performance
188188
optimizer:__choice__, Value: 'AdamOptimizer'
189189
scaler:__choice__, Value: 'StandardScaler'
190190
trainer:__choice__, Value: 'StandardTrainer'
191-
, ta_runs=0, ta_time_used=0.0, wallclock_time=0.0010449886322021484, budget=0), TrajEntry(train_perf=4.9408606026215836, incumbent_id=1, incumbent=Configuration:
191+
, ta_runs=0, ta_time_used=0.0, wallclock_time=0.0011219978332519531, budget=0), TrajEntry(train_perf=4.9408606026215836, incumbent_id=1, incumbent=Configuration:
192192
data_loader:batch_size, Value: 64
193193
encoder:__choice__, Value: 'OneHotEncoder'
194194
feature_preprocessor:__choice__, Value: 'NoFeaturePreprocessor'
@@ -219,7 +219,7 @@ Print the final ensemble performance
219219
optimizer:__choice__, Value: 'AdamOptimizer'
220220
scaler:__choice__, Value: 'StandardScaler'
221221
trainer:__choice__, Value: 'StandardTrainer'
222-
, ta_runs=1, ta_time_used=1.316178560256958, wallclock_time=2.6122844219207764, budget=5.555555555555555), TrajEntry(train_perf=1.0, incumbent_id=2, incumbent=Configuration:
222+
, ta_runs=1, ta_time_used=1.2900512218475342, wallclock_time=2.4858930110931396, budget=5.555555555555555), TrajEntry(train_perf=1.0, incumbent_id=2, incumbent=Configuration:
223223
data_loader:batch_size, Value: 165
224224
encoder:__choice__, Value: 'NoEncoder'
225225
feature_preprocessor:TruncatedSVD:target_dim, Value: 5
@@ -245,7 +245,7 @@ Print the final ensemble performance
245245
optimizer:__choice__, Value: 'SGDOptimizer'
246246
scaler:__choice__, Value: 'StandardScaler'
247247
trainer:__choice__, Value: 'StandardTrainer'
248-
, ta_runs=5, ta_time_used=9.937701940536499, wallclock_time=16.307201385498047, budget=5.555555555555555), TrajEntry(train_perf=1.0312830279463878, incumbent_id=3, incumbent=Configuration:
248+
, ta_runs=5, ta_time_used=9.524490118026733, wallclock_time=15.669584274291992, budget=5.555555555555555), TrajEntry(train_perf=0.9809322573013144, incumbent_id=3, incumbent=Configuration:
249249
data_loader:batch_size, Value: 56
250250
encoder:__choice__, Value: 'OneHotEncoder'
251251
feature_preprocessor:Nystroem:kernel, Value: 'cosine'
@@ -283,7 +283,7 @@ Print the final ensemble performance
283283
scaler:__choice__, Value: 'NoScaler'
284284
trainer:MixUpTrainer:alpha, Value: 0.19515049217885105
285285
trainer:__choice__, Value: 'MixUpTrainer'
286-
, ta_runs=10, ta_time_used=23.166086196899414, wallclock_time=34.92461705207825, budget=16.666666666666664)]
286+
, ta_runs=10, ta_time_used=22.451351165771484, wallclock_time=33.967867851257324, budget=16.666666666666664)]
287287
{'r2': 0.9153034112747092}
288288
| | Preprocessing | Estimator | Weight |
289289
|---:|:----------------|:------------|---------:|
@@ -296,7 +296,7 @@ Print the final ensemble performance
296296
297297
.. rst-class:: sphx-glr-timing
298298

299-
**Total running time of the script:** ( 5 minutes 38.063 seconds)
299+
**Total running time of the script:** ( 5 minutes 39.929 seconds)
300300

301301

302302
.. _sphx_glr_download_examples_20_basics_example_tabular_regression.py:

‎development/_sources/examples/20_basics/sg_execution_times.rst.txt‎

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -5,12 +5,12 @@
55

66
Computation times
77
=================
8-
**11:08.456** total execution time for **examples_20_basics** files:
8+
**11:11.780** total execution time for **examples_20_basics** files:
99

1010
+--------------------------------------------------------------------------------------------------------------+-----------+--------+
11-
| :ref:`sphx_glr_examples_20_basics_example_tabular_regression.py` (``example_tabular_regression.py``) | 05:38.063 | 0.0 MB |
11+
| :ref:`sphx_glr_examples_20_basics_example_tabular_regression.py` (``example_tabular_regression.py``) | 05:39.929 | 0.0 MB |
1212
+--------------------------------------------------------------------------------------------------------------+-----------+--------+
13-
| :ref:`sphx_glr_examples_20_basics_example_tabular_classification.py` (``example_tabular_classification.py``) | 05:24.604 | 0.0 MB |
13+
| :ref:`sphx_glr_examples_20_basics_example_tabular_classification.py` (``example_tabular_classification.py``) | 05:25.351 | 0.0 MB |
1414
+--------------------------------------------------------------------------------------------------------------+-----------+--------+
15-
| :ref:`sphx_glr_examples_20_basics_example_image_classification.py` (``example_image_classification.py``) | 00:05.789 | 0.0 MB |
15+
| :ref:`sphx_glr_examples_20_basics_example_image_classification.py` (``example_image_classification.py``) | 00:06.500 | 0.0 MB |
1616
+--------------------------------------------------------------------------------------------------------------+-----------+--------+

0 commit comments

Comments
 (0)