)
pytorch_step = EstimatorStep(
name="PyTorch_Train",
estimator=pytorch_est,
estimator_entry_script_arguments=["--data_dir", output],
inputs=[output],
compute_target=cpu_cluster,
allow_reuse=True,
)
pipeline = Pipeline(workspace=ws, steps=[preprocessing_step, pytorch_step])
run = Experiment(ws, args.experiment).submit(pipeline)
run.wait_for_completion(show_output=True)
preprocessing_run = next(step for step in run.get_steps()
if step.name == "Preprocessing_Train")
pytorch_run = next(step for step in run.get_steps()
if step.name == "PyTorch_Train")
preprocessing_model = preprocessing_run.register_model(
model_name='sklearn_preprocessing',
model_path='outputs/preprocessing.joblib')
print(preprocessing_model.name,
preprocessing_model.id,
preprocessing_model.version,
sep='\t')
pytorch_model = pytorch_run.register_model(model_name='autoencoder',
model_path='outputs/model.pt')
print(pytorch_model.name, pytorch_model.id, pytorch_model.version, sep='\t')
from azureml.pipeline.core import Pipeline
# Build the pipeline
pipeline1 = Pipeline(workspace=ws, steps=[compareModels])
run = Run.get_context()
# Envoyer le pipeline
from azureml.core import Experiment
# Submit the pipeline to be run
pipeline_run1 = Experiment(ws, 'Compare_Models_Exp').submit(pipeline1)
pipeline_run1.wait_for_completion(False, 0, True)
#Récupération du run id de la step 1 sur 1 de la pipeline
step1_runid = [e for e in pipeline_run1.get_steps()][0].id
#création du path complet dans le blob storage
path_model_datastore = "azureml/" + step1_runid + "/outputs/"
#download en local du modele
def_data_store.download(target_path='datastore', prefix=path_model_datastore)
#chargement dans la workspace a partir du modele download en local
model = Model.register(workspace=ws,
model_path='datastore/' + path_model_datastore +
model_pklname + '.pkl',
model_name=model_basename)
#Create environment file
from azureml.core.conda_dependencies import CondaDependencies
myenv = CondaDependencies()
myenv.add_conda_package("scikit-learn")
allow_reuse=True
)
# Build pipeline
pipeline_steps = [
dataprep_step,
train_step
]
pipeline = Pipeline(workspace=workspace, steps=[pipeline_steps])
# Run pipeline
run = Experiment(workspace=workspace, name='gensim_lda-pipeline').submit(pipeline)
run.wait_for_completion(show_output=True)
# Get training step
run_train_step = [s for s in run.get_steps() if s.name == 'train.py'][0]
print(run_train_step.get_metrics())
print(run_train_step.get_file_names())
# Register model
model = run_train_step.register_model(model_name='gensim_lda', model_path='outputs')
print(model.name, model.id, model.version, sep='\t')
