def make_predictions(): #We will make predictions for those experiments, which are already finished, but don't have the csv file with predictions
experiments = projects.Project(Path(__file__).parent.parent).experiments()
for exp in experiments:
res_file = os.path.join(exp.path, exp.name() + '_result.csv')
if exp.isCompleted() and not os.path.exists(res_file):
preds = generic.parse(exp.path).predictions('test')
preds.dump(res_file)
def __init__(self,path,input_columns,output_columns,ctypes={},input_groups={},output_groups={}):
self.cfg=generic.parse(path)
path=os.path.dirname(os.path.dirname(path))
self.cfg._projectDir=path
self.input_columns=input_columns
self.output_columns=output_columns
self.image_path=[]
self.ctypes=ctypes
self.path=path
self.input_groups=input_groups
self.output_groups=output_groups
def translate(sentence: str):
file_path = os.path.join(context.get_current_project_data_path(),
"rus.vocab")
vocabulary = utils.load(file_path)
preds = generic.parse('eng_to_ru').predictions(
builtin_datasets.from_array([sentence], ['']))
for item in preds:
rootItem = item.rootItem()
sentence = ''
for indices in item.prediction:
sentence = sentence + " " + vocabulary.i2w[np.argmax(indices)]
print(rootItem.x + " " + sentence)
def parse_config(self):
extra = None
if self.project is not None:
if os.path.exists(self.project.commonPath()):
extra = self.project.commonPath()
if os.path.exists(self.getConfigYamlPath()):
cfg = generic.parse(self.getConfigYamlPath(), extra)
else:
cfg = generic.parse(self.getConfigYamlConcretePath(), extra)
cfg.gpus = self.gpus
if self.allowResume:
cfg.setAllowResume(self.allowResume)
if self.project is not None:
if os.path.exists(self.project.modulesPath()):
for m in os.listdir(self.project.modulesPath()):
if ".py" in m:
cfg.imports.append(m[0:m.index(".py")])
if os.path.exists(self.project.dataPath()):
cfg.datasets_path = self.project.dataPath()
return cfg
def make_predictions():
experiments = projects.Project(Path(__file__).parent.parent).experiments()
for exp in experiments:
if exp.isCompleted():
file_path=os.path.join(context.get_current_project_data_path(),"rus.vocab")
vocabulary=utils.load(file_path)
preds = generic.parse(exp.path).predictions('test')
for item in preds:
rootItem = item.rootItem()
sentence = ''
for indices in item.prediction:
sentence = sentence + " " + vocabulary.i2w[np.argmax(indices)]
print(rootItem.x + " " + sentence)
def blend(self, ds, w=0.5) -> DataSet:
if isinstance(ds, str):
from musket_core import generic
return self.blend(generic.parse(ds).predictions(self.parent.name))
return self._inner_blend(ds, w)
def make_predictions():
preds = generic.parse("questions1").predictions('test')
preds.dump('predictions.csv')
def make_predictions():
preds = generic.parse("simpleCnn").predictions('Test')
preds.dump('predictions_cnn.csv')
preds = generic.parse("simpleRnn").predictions('Test')
preds.dump('predictions_rnn.csv')
