The documentation doesn't appear to document how to use .options(...) in any of the following

remote_function.options(...).remote()
ActorClass.options(...).remote()

Less important, but we should also probably introduce

actor.method.options(...).remote()

Hi,
I'm new to tpot but I got this error. I understand that score function can take strings, but I got the following error when using TPOTClassifier.

---

ValueError Traceback (most recent call last)
in
----> 1 tpot.score(X_test, y_test)

~/miniconda3/envs/ml

is it Grid Search can solve CASH problems with NNI , it seems that it is usually used for hyper-parameters optimization, have you guys have finished some revision for Grid Search for solving CASH problems.

about Cash problems can refer to ：microsoft/nni#1178

i want to know whether autosklearn support multilabel task, thank you!

As a user/researcher I would like to easily be able to find information on what kind of algorithm stands behind optuna, to be able to make informed decision on using it in my work. Currently main page says only that it does "Bayesian optimization".

As a user I would like to see detailed documentation about the algorithm behind the Study.optimize method, when I check the documentation page for

According to issue #204 the number of layers could be tuned by a for-loop

model = ...
num_layers = <result of randint>
for _ in range(num_layers):
        model.add(Dense({{choice([np.power(2,5),np.power(2,9),np.power(2,11)])}}))
        model.add(Activation({{choice(['tanh','relu', 'sigmoid'])}}))
        model.add(Dropout({{uniform(0, 1)}}))

But the layers added in the for-

pedantic, I know, but README.md has a typo in the Categorical Ensembling section where it refers to markets in the first sentence, not models.

Sklearn has a TON of estimators, metrics, and cv iterators that could trivially be added to the xcessiv.presets package. I'm a bit focused on other issues to bother adding them all.

Anyone who can help add to the list can easily do so.

Adding preset estimators/metrics/cvs is very easy. There's literally no need to understand how the rest of Xcessiv works, just take a look and copy the patt

• Unable to supply validation_data to a Keras CVExperiment via model_extra_params[“fit”]
• This is because HyperparameterHunter automatically sets validation_data to be the OOF data produced by the cross validation scheme
• I can imagine this would be unexpected behavior, so I’d love to hear any thoughts on how to clear this up

Note

• This issue (along with several others) was ori

Currently the hparams are logged as text in tensorboard. Could we change this to start using the add_hparams() function in SummaryWritter. This will allow some additional nice views of the hyperparameters such as parallel coordinates view.

I think this would be a relatively easy change and I can help out with it if needed.

[question] [feature request]

Am enjoying r-b-z. Wanting to make envs, or wanting to modify them.

I know you can define a custom env as described in link 1 and in link 2.

The problem with link 1 is that you're limited in the number of paramet

If enter_data() is called with the same train_path twice in a row and the data itself hasn't changed, a new Dataset does not need to be created.

We should add a column which stores some kind of hash of the actual data. When a Dataset would be created, if the metadata and data hash are exactly the same as an existing Dataset, nothing should be added to the ModelHub database and the existing

Expected Results

List of cost of all configs, or run if more then one run per config

Actual Results

Don't know where it is

Versions

0.11.1

I can't see anywhere in the documentation exactly what min and max budget mean. At first I saw examples that made it look like they are the number of epochs, but given the default values are 0.01 and 1, respectively, that doesn't make much sense.

Are they arbitrary, but it is simply the ratio of the two that determines the step sizes e.g. during successive halving?

When running the firststeps example I get several deprecation warnings. The environment is a fresh virtual conda environment on Win 10 x64 with the following package versions: packages.txt

First the tensorflow warnings:

WARNING:tensorflow:From [envpat

Hello all,

I started writing an sqlite db implementation for Orion. It currently is in working state but I need to add tests, improve comments etc. When I do those I will do a pull request.

It is not an issue that I am reporting, more an attempt to start a conversation around developing new db implementations and the design decisions that have been made and make it harder to implement a db c

Ability to add arguments to Quantized distributions to make them not only perfect integers, but rather, also add an interval such as interval=0.1 to have more precision on the quantification, and perhaps even a fixed phase to start the quantization, such as saying also anchor=0.05.

So a Quantized(Uniform(0, 1), interval=0.1, anchor=0.05) would generate something like `[0.05, 0.15, 0.25,

Hi,

Thanks for sharing the code.

Is there ant simple "getting started" document on setting up like a basic EI or KG with your tool.
1- I have seen the main.py in the "example" folder, it is good but it is for a very complete case, not for simple sampling mode.
2- I have tried the MOE getting started document (available in the "doc" folder) but as the function names and procedures are chan

opt.ea.lambda | [numeric{1}]
For opt.ea = "ea". Number of children generated in each generation. Default is 1.

should this be an integer(1)?

We should focus over the new documentation of BTB and mainly on the following points:

• Update the Tuners section, including how to contribute a new tuner.
• Update the Hyperparams section.
• Create new section explaining how to create and use the Tunable.
• Create a new README where BTBSession is used, which I belive should be the main focus of the library at the moment.