2f5d6169d3
example update ( #359 )
...
update some examples for consistencies with others.
2021-12-25 16:13:39 -08:00
b2900f4b22
fixing custom metric ( #357 )
...
* fixing the error for custom metric
2021-12-24 16:23:09 -05:00
c6c0c29769
Simplify lgbm example ( #358 )
...
* simplify lgbm examples
* provide link to lgbm example script.
* simply lgbm example in the example script.
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
2021-12-23 23:05:14 -08:00
dcfd218108
Fixing the bug in custom metric ( #356 )
...
* fixing the bug for custom metric
2021-12-23 18:44:53 -05:00
300f286667
azureml + ray ( #344 )
...
* examples and documentation about how to use azureml + ray
2021-12-23 13:37:07 -08:00
baa0359324
doc update ( #352 )
...
* custom splitter
* NLP
* version number
2021-12-22 14:35:13 -08:00
0b25e89f29
reproducibility for random sampling ( #349 )
...
* reproducibility for random sampling #236
* doc update
2021-12-22 12:12:25 -08:00
ee3162e232
Adding the NLP task summarization ( #346 )
...
* Add test_autohf_summarization.py
* adding seq2seq
* Update flaml/nlp/huggingface/trainer.py
* rouge metrics
Co-authored-by: XinZofStevens <xzhao4346@gmail.com>
Co-authored-by: JinzhuoWu <wujinzhuo0105@gmail.com>
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
2021-12-20 14:19:32 -08:00
a340dbc554
url update for doc edit
2021-12-18 15:00:54 -08:00
f3fc35c438
version update
2021-12-16 18:38:19 -08:00
efd85b4c86
Deploy a new doc website ( #338 )
...
A new documentation website. And:
* add actions for doc
* update docstr
* installation instructions for doc dev
* unify README and Getting Started
* rename notebook
* doc about best_model_for_estimator #340
* docstr for keep_search_state #340
* DNN
Co-authored-by: Qingyun Wu <qingyun.wu@psu.edu>
Co-authored-by: Z.sk <shaokunzhang@psu.edu>
2021-12-16 17:11:33 -08:00
671ccbbe3f
support for customized splitters ( #333 )
...
* add support for customized splitters
* use the param split_type for feeding generators
* use single API for customized splitter and add test
* when task==TS_FORCAST, always set shuffle=False
* update docstr
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
2021-12-16 16:13:04 -08:00
7b24662dca
Makes the evaluation_function could receive the incumbent best result as input in Tune ( #339 )
...
* update tune function
* pass incumbent result to the training function
* Update test/tune/test_record_incumbent.py
* Update flaml/searcher/search_thread.py
* Update flaml/searcher/blendsearch.py
* Update flaml/tune/tune.py
* add constant variable
Co-authored-by: 张少坤 <zhangshaokun@fuzhi.ai>
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
2021-12-15 21:12:47 -08:00
228b00cbee
use cfo in tune.run if bs is not installed ( #334 )
...
Use cfo in tune.run if bs is not installed and issue a warning.
2021-12-12 21:10:06 -08:00
434586e2e2
train at least one iter when not trained ( #336 )
...
* train at least one iter when not trained
* bump version to 0.9.1
2021-12-12 20:05:18 -08:00
1a3e01c352
adding HF metrics ( #335 )
...
* adding nlp metrics
* fix ndcg
2021-12-10 12:32:49 -05:00
b773e2898f
Update flaml_pytorch_cifar10.ipynb
...
training_function -> train_cifar
2021-12-07 15:17:15 -08:00
dd60dbc5eb
rename training_function ( #327 )
...
* rename training_function
* add docstr
* update docstr
* update docstr and comments
Co-authored-by: Qingyun Wu <qxw5138@psu.edu>
2021-12-06 17:03:43 -05:00
3111084c07
add __init__.py in nlp
2021-12-06 09:15:39 -08:00
17b17d084f
tune api for schedulers ( #322 )
...
* revise api and tests
* rename prune_attr
* update finetune notebook
* add scheduler test and notebook
* update tune api for scheduler
* remove scheduler notebook
* Update flaml/tune/tune.py
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
* docstr
* fix imports
* clear notebook output
* fix ray import
* Update flaml/tune/tune.py
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
* improve docstr
* Update flaml/searcher/blendsearch.py
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
* remove redundant import
Co-authored-by: Qingyun Wu <qxw5138@psu.edu>
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
2021-12-04 21:52:20 -05:00
7d269435ae
add save_best_config()
2021-12-04 16:29:52 -08:00
54d303a95a
bug fix in confg2params ( #323 )
...
* bug fix in confg2params
* set the task property before config2params
2021-12-03 19:37:49 -08:00
18230ed22f
pred_time_limit clarification and logging ( #319 )
...
* pred_time_limit clarification
* log prediction time
* handle ChunkedEncodingError in test
2021-12-03 16:02:00 -08:00
fb59bb9928
adding TODOs for NLP module, so students can implement other tasks easier ( #321 )
...
* fixing ray pickle bug, skipping macosx bug, completing code for seqregression
* catching connectionerror
* ading TODOs for NLP module
2021-12-03 12:45:16 -05:00
c57954fbbd
include default value in rf search space ( #317 )
...
* include default value in rf search space
* init _mem_per_iter with -1
* bump version to 0.8.2
* docstr for search space's arguments
2021-12-03 09:15:21 -08:00
1545d5a6d2
skip cv preparation if eval_method is holdout ( #314 )
...
* skip cv preparation if eval_method is holdout
* bump version to 0.8.1
2021-11-28 11:18:55 -08:00
2f25a87d98
Code quality improvement based on #275 ( #313 )
...
* simplify & restructure
Co-authored-by: Albern S <62778698+albernsrya@users.noreply.github.com>
2021-11-28 10:14:25 -08:00
63f402b29e
fixing config2params for transformersestimator
2021-11-26 21:28:38 -08:00
b0ef3b7995
Add conda forge minimal test ( #309 )
...
* add conda forge minimal test, create pytest markers
2021-11-25 13:58:20 -08:00
5dc948da18
Update test_regression.py ( #306 )
...
* Update test_regression.py
There is a another way for mutioutput-model-trian.
RegressorChain is more adapting to the targets which are relavant.
2021-11-25 08:18:22 -08:00
fd136b02d1
bug fix for TransformerEstimator ( #293 )
...
* fix checkpoint naming + trial id for non-ray mode, fix the bug in running test mode, delete all the checkpoints in non-ray mode
* finished testing for checkpoint naming, delete checkpoint, ray, max iter = 1
* adding predict_proba, address PR 293's comments
close #293 #291
2021-11-23 11:26:39 -08:00
85e21864ce
test -> val; docstr ( #300 )
...
* rename test -> val in custom metric function
* add an example in docstr
resolve #299
2021-11-22 22:17:29 -08:00
ea6d28d7bd
add max_depth to xgboost search space ( #282 )
...
* add max_depth to xgboost search space
* notebook update
* two learners for xgboost (max_depth or max_leaves)
2021-11-22 21:17:48 -08:00
d937b03e42
multioutput regression ( #292 )
...
* make AutoML inherit sklearn.base.BaseEstimator such that it can be wrapped in sklearn.multioutput.MultiOutputRegressor for multi-output regression.
* moved and simplified preprocessing code in AutoML.predictI() to _preprocess()
2021-11-22 06:59:42 -08:00
00da79a90b
empty search space ( #295 )
...
fix the error when an empty dictionary is passed to BlendSearch as the search space.
2021-11-20 20:05:28 -08:00
49f9e9f86b
add warmstart test ( #298 )
...
* add warmstart test
* remove redundancy
* add more types of hps
* revise comments
* simplify name
* reduce redundancy
2021-11-20 20:23:54 -05:00
db1fb9b47b
datetime feature engineering ( #285 )
...
resolve #284
When transforming test data, keep a derived column as long as it is kept in the training data.
2021-11-18 11:19:53 -08:00
72caa2172d
model_history, ITER_HP, settings in AutoML(), checkpoint bug fix ( #283 )
...
if save_best_model_per_estimator is False and retrain_final is True, unfit the model after evaluation in HPO.
retrain if using ray.
update ITER_HP in config after a trial is finished.
change prophet logging level.
example and notebook update.
allow settings to be passed to AutoML constructor. Are you planning to add multi-output-regression capability to FLAML #192 Is multi-tasking allowed? #277 can pass the auotml setting to the constructor instead of requiring a derived class.
remove model_history.
checkpoint bug fix.
* model_history meaning save_best_model_per_estimator
* ITER_HP
* example update
* prophet logging level
* comment update in forecast notebook
* print format improvement
* allow settings to be passed to AutoML constructor
* checkpoint bug fix
* time limit for autohf regression test
* skip slow test on macos
* cleanup before del
2021-11-18 09:39:45 -08:00
e9551de3cc
add best_loss_per_estimator
2021-11-17 22:43:20 -08:00
42de3075e9
Make NLP tasks available from AutoML.fit() ( #210 )
...
Sequence classification and regression: "seq-classification" and "seq-regression"
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
2021-11-16 11:06:20 -08:00
59083fbdcb
example update ( #281 )
...
* example update
* bump version to 0.7.2
* notebook update
2021-11-12 22:29:33 -08:00
92ebd1f7f9
when max_iter=1, skip search only if retrain_final ( #280 )
...
* when max_iter=1, skip search only if retrain_final
* remove nlp
redesign in #210
* minor change in readme example
2021-11-09 21:51:23 -08:00
5b0932e442
Unify regression and classification for XGBoost ( #276 )
...
* scikit-learn API for XGBoostRegressor
2021-11-09 21:23:54 -08:00
3f09c694a3
Merge pull request #274 from microsoft/docstr
...
update docstr
2021-11-08 09:59:23 +08:00
5b68f556dc
bump version to 0.7.1
2021-11-07 08:05:13 -08:00
0c7bf7219f
Merge branch 'main' into docstr
2021-11-06 21:58:26 -07:00
62a31704ee
default to cfo for single estimator ( #273 )
...
* default to cfo for single estimator
* use bs for parallel tuning
* comment about overhead
2021-11-06 21:58:05 -07:00
09e75c9b9c
fix onlineaml docstr
2021-11-07 02:46:27 +00:00
03bc62363f
add periods
2021-11-06 12:44:10 -07:00
c4d5986ee8
no retraining when max_iter=0 and not retrain_full
2021-11-06 11:37:57 -07:00
Chi Wang