Manage

Manage

To ensure that your submitted code identity is correctly recognized by Gitee, please execute the following command.

git config --global user.name userName git config --global user.email userEmail

When using the SSH protocol for the first time to clone or push code, follow the prompts below to complete the SSH configuration.

1 Generate RSA keys.

2 Obtain the content of the RSA public key and configure it in SSH Public Keys

To use SVN on Gitee, please visit the usage guide

When using the HTTPS protocol, the command line will prompt for account and password verification as follows. For security reasons, Gitee recommends configure and use personal access tokens instead of login passwords for cloning, pushing, and other operations.

Username for 'https://gitee.com': userName

Password for 'https://userName@gitee.com': # Private Token

File empty ...

MLproject

README.rst

conda.yaml

search_gpyopt.py

search_hyperopt.py

search_random.py

train.py

Hyperparameter Tuning Example

Example of how to do hyperparameter tuning with MLflow and some popular optimization libraries.

This example tries to optimize the RMSE metric of a Keras deep learning model on a wine quality dataset. The Keras model is fitted by the train entry point and has two hyperparameters that we try to optimize: learning-rate and momentum. The input dataset is split into three parts: training, validation, and test. The training dataset is used to fit the model and the validation dataset is used to select the best hyperparameter values, and the test set is used to evaluate expected performance and to verify that we did not overfit on the particular training and validation combination. All three metrics are logged with MLflow and you can use the MLflow UI to inspect how they vary between different hyperparameter values.

examples/hyperparam/MLproject has 4 targets:

• train train simple deep learning model on the wine-quality dataset from our tutorial. It has 2 tunable hyperparameters: learning-rate and momentum. Contains examples of how Keras callbacks can be used for MLflow integration.
• random perform simple random search over the parameter space.
• gpyopt use Hyperopt to optimize hyperparameters.

All the hyperparameter targets take an optional experiment ID for training runs. If provided, training runs are logged under this experiment ID. This organizes the runs so that it is easy to view individual training runs and the hyperparameter runs separately.

Running this Example

You can run any of the targets as a standard MLflow run.

mlflow experiments create individual_runs

Creates experiment for individual runs and return its experiment ID.

mlflow experiments create hyper_param_runs

Creates an experiment for hyperparam runs and return its experiment ID.

mlflow run -e train --experiment-id <individual_runs_experiment_id> examples/hyperparam

Runs the Keras deep learning training with default parameters and log it in experiment 1.

mlflow run -e random --experiment-id <hyperparam_experiment_id> -P \
 training_experiment_id=<individual_runs_experiment_id> examples/hyperparam

mlflow run -e gpyopt --experiment-id <hyperparam_experiment_id> -P \
 training_experiment_id=<individual_runs_experiment_id> examples/hyperparam

mlflow run -e hyperopt --experiment-id <hyperparam_experiment_id> -P \
 training_experiment_id=<individual_runs_experiment_id> examples/hyperparam

Runs the hyperparameter tuning with either random search or GpyOpt or Hyperopt and log the results under hyperparam_experiment_id.

You can compare these results by using mlflow ui.

Report success

Type

Reason

取消

提交