def __call__(self, func, *args, **kwargs):
if isinstance(func, FilterInterface):
return DataProcessor(self._data.loc[func.apply(self._data)])
else:
ret_value = apply_func(self._data, func, *args, **kwargs)
if not isinstance(ret_value, type(self._data)):
raise TypeError(
f'Call to DataProcessor should return type {type(self._data)} but returned {type(ret_value)}')
return DataProcessor(ret_value)
"id":
row.get(id_labels_column),
"text":
row.get(text_column),
"lang":
row.get(lang_column),
"concept_id":
row.get(concept_id_column),
"precision":
row.get(precision_column)
}]).dict()
def train_model(client):
client.model.train(model_id)
return str(client.model.wait_training(model_id))
cl_client = get_client(get_recipe_config())
init_model(cl_client)
apply_func(upload_concepts,
client=cl_client,
input_dataset="concepts_input_dataset",
output_dataset="concepts_output_dataset")
apply_func(upload_labels,
client=cl_client,
input_dataset="labels_input_dataset",
output_dataset="labels_output_dataset")
train_model(cl_client)
for k, v in row.items() if k.startswith(variation_prefix)
}
return client.faq.create_questions(model_id,
questions=[{
"id":
row.get(questions_id_column),
"variations":
variations,
"answer_id":
row.get(answer_id_column)
}]).dict()
def train_model(client):
client.model.train(model_id)
return str(client.model.wait_training(model_id))
cl_client = get_client(get_recipe_config())
init_model(cl_client)
apply_func(upload_answers,
client=cl_client,
input_dataset="answers_input_dataset",
output_dataset="answers_output_dataset")
apply_func(upload_questions,
client=cl_client,
input_dataset="questions_input_dataset",
output_dataset="questions_output_dataset")
train_model(cl_client)
from dataiku.customrecipe import get_recipe_config
from utils import apply_func
url_column = get_recipe_config().get("url_column")
def call_api(client, row):
return client.html.extract_article(row.get(url_column)).dict()
apply_func(call_api)
except ModelNotFound:
client.model.create(model_id, "clf")
def call_api(client, row):
return client.classifier.create_documents(model_id,
documents=[{
"text":
row.get(text_column),
"lang":
row.get(lang_column),
"class_id":
row.get(class_id_column),
"id":
row.get(id_column)
}]).dict()
def train_model(client):
client.model.train(model_id,
model_type=model_type,
model_config={"train_ratio": train_ratio})
return str(cl_client.model.wait_training(model_id))
cl_client = get_client(get_recipe_config())
init_model(cl_client)
apply_func(call_api, client=cl_client)
train_model(cl_client)
