def main(config="../../config.yaml", namespace=""):
# obtain config
if isinstance(config, str):
config = load_job_config(config)
backend = config.backend
work_mode = config.work_mode
lr_param = {
"name": "hetero_lr_0",
"penalty": "L2",
"optimizer": "rmsprop",
"tol": 0.0001,
"alpha": 0.01,
"max_iter": 10,
"early_stop": "diff",
"batch_size": -1,
"learning_rate": 0.15,
"init_param": {
"init_method": "zeros"
},
"cv_param": {
"n_splits": 3,
"shuffle": False,
"random_seed": 103,
"need_cv": True
}
}
pipeline = common_tools.make_normal_dsl(config, namespace, lr_param, is_cv=True)
# fit model
pipeline.fit(backend=backend, work_mode=work_mode)
# query component summary
common_tools.prettify(pipeline.get_component("hetero_lr_0").get_summary())
def main(config="../../config.yaml", namespace=""):
# obtain config
if isinstance(config, str):
config = load_job_config(config)
lr_param = {
"name": "hetero_lr_0",
"penalty": "L2",
"optimizer": "nesterov_momentum_sgd",
"tol": 1e-05,
"alpha": 0.0001,
"max_iter": 1,
"early_stop": "diff",
"multi_class": "ovr",
"batch_size": -1,
"learning_rate": 0.15,
"init_param": {
"init_method": "zeros"
}
}
pipeline = common_tools.make_normal_dsl(config, namespace, lr_param, is_ovr=True)
# fit model
pipeline.fit()
# query component summary
common_tools.prettify(pipeline.get_component("hetero_lr_0").get_summary())
def main(config="../../config.yaml", namespace=""):
# obtain config
if isinstance(config, str):
config = load_job_config(config)
lr_param = {
"name": "hetero_lr_0",
"penalty": "L2",
"optimizer": "rmsprop",
"tol": 0.0001,
"alpha": 0.01,
"max_iter": 30,
"early_stop": "diff",
"batch_size": 320,
"batch_strategy": "random",
"learning_rate": 0.15,
"init_param": {
"init_method": "zeros"
},
"sqn_param": {
"update_interval_L": 3,
"memory_M": 5,
"sample_size": 5000,
"random_seed": None
},
"cv_param": {
"n_splits": 5,
"shuffle": False,
"random_seed": 103,
"need_cv": False
},
"callback_param": {
"callbacks": ["ModelCheckpoint"],
"save_freq": "epoch"
}
}
pipeline = common_tools.make_normal_dsl(config, namespace, lr_param)
# dsl_json = predict_pipeline.get_predict_dsl()
# conf_json = predict_pipeline.get_predict_conf()
# import json
# json.dump(dsl_json, open('./hetero-lr-normal-predict-dsl.json', 'w'), indent=4)
# json.dump(conf_json, open('./hetero-lr-normal-predict-conf.json', 'w'), indent=4)
# fit model
pipeline.fit()
# query component summary
common_tools.prettify(pipeline.get_component("hetero_lr_0").get_summary())
common_tools.prettify(pipeline.get_component("evaluation_0").get_summary())
def main(config="../../config.yaml", namespace=""):
# obtain config
if isinstance(config, str):
config = load_job_config(config)
backend = config.backend
work_mode = config.work_mode
lr_param = {
"name": "hetero_lr_0",
"penalty": "L2",
"optimizer": "rmsprop",
"tol": 0.0001,
"alpha": 0.01,
"max_iter": 30,
"early_stop": "diff",
"batch_size": -1,
"learning_rate": 0.15,
"validation_freqs": 1,
"early_stopping_rounds": 3,
"metrics": [],
"use_first_metric_only": False,
"init_param": {
"init_method": "zeros"
},
"sqn_param": {
"update_interval_L": 3,
"memory_M": 5,
"sample_size": 5000,
"random_seed": None
},
"cv_param": {
"n_splits": 5,
"shuffle": False,
"random_seed": 103,
"need_cv": False
}
}
pipeline = common_tools.make_normal_dsl(config,
namespace,
lr_param,
has_validate=True)
# fit model
job_parameters = JobParameters(backend=backend, work_mode=work_mode)
pipeline.fit(job_parameters)
# query component summary
common_tools.prettify(pipeline.get_component("hetero_lr_0").get_summary())
def main(config="../../config.yaml", namespace=""):
# obtain config
if isinstance(config, str):
config = load_job_config(config)
backend = config.backend
work_mode = config.work_mode
lr_param = {
"name": "hetero_lr_0",
"penalty": "L2",
"optimizer": "sqn",
"tol": 0.0001,
"alpha": 1e-05,
"max_iter": 30,
"early_stop": "diff",
"batch_size": 5000,
"learning_rate": 0.15,
"decay": 0.3,
"decay_sqrt": True,
"init_param": {
"init_method": "zeros"
},
"sqn_param": {
"update_interval_L": 3,
"memory_M": 5,
"sample_size": 5000,
"random_seed": None
},
"cv_param": {
"n_splits": 5,
"shuffle": False,
"random_seed": 103,
"need_cv": False
}
}
pipeline = common_tools.make_normal_dsl(config,
namespace,
lr_param,
is_dense=False)
# fit model
pipeline.fit(backend=backend, work_mode=work_mode)
# query component summary
common_tools.prettify(pipeline.get_component("hetero_lr_0").get_summary())
common_tools.prettify(pipeline.get_component("evaluation_0").get_summary())
def main(config="../../config.yaml", namespace=""):
# obtain config
if isinstance(config, str):
config = load_job_config(config)
lr_param = {
"name": "hetero_lr_0",
"penalty": "L2",
"optimizer": "rmsprop",
"tol": 0.0001,
"alpha": 0.01,
"max_iter": 30,
"callback_param": {
"callbacks": ["EarlyStopping"],
"validation_freqs": 3,
"early_stopping_rounds": 3
},
"early_stop": "diff",
"batch_size": -1,
"learning_rate": 0.15,
"init_param": {
"init_method": "zeros",
"fit_intercept": True
},
"encrypt_param": {
"key_length": 2048
},
"cv_param": {
"n_splits": 5,
"shuffle": False,
"random_seed": 103,
"need_cv": False
}
}
pipeline = common_tools.make_normal_dsl(config,
namespace,
lr_param,
has_validate=True)
# fit model
pipeline.fit()
# query component summary
common_tools.prettify(pipeline.get_component("evaluation_0").get_summary())
def main(config="../../config.yaml", namespace=""):
# obtain config
if isinstance(config, str):
config = load_job_config(config)
lr_param = {
"name": "hetero_lr_0",
"penalty": "L2",
"optimizer": "nesterov_momentum_sgd",
"tol": 0.0001,
"alpha": 0.01,
"max_iter": 30,
"early_stop": "weight_diff",
"batch_size": -1,
"learning_rate": 0.15,
"init_param": {
"init_method": "zeros"
},
"sqn_param": {
"update_interval_L": 3,
"memory_M": 5,
"sample_size": 5000,
"random_seed": None
},
"cv_param": {
"n_splits": 5,
"shuffle": False,
"random_seed": 103,
"need_cv": False
}
}
pipeline = common_tools.make_normal_dsl(config,
namespace,
lr_param,
is_multi_host=True)
# fit model
pipeline.fit()
# query component summary
common_tools.prettify(pipeline.get_component("hetero_lr_0").get_summary())
