Split sacred and MLflow to separate files
This commit is contained in:
parent
72c11b869f
commit
b7556dedb0
80
lab07_sacred.py
Normal file
80
lab07_sacred.py
Normal file
@ -0,0 +1,80 @@
|
|||||||
|
from datetime import datetime
|
||||||
|
|
||||||
|
import pandas as pd
|
||||||
|
from sacred import Experiment
|
||||||
|
from sacred.observers import FileStorageObserver
|
||||||
|
from sklearn.metrics import mean_squared_error
|
||||||
|
from sklearn.model_selection import train_test_split
|
||||||
|
from tensorflow.keras.callbacks import EarlyStopping
|
||||||
|
from tensorflow.keras.layers import Dense, Dropout
|
||||||
|
from tensorflow.keras.models import Sequential
|
||||||
|
|
||||||
|
ex = Experiment("file_observer", interactive=False, save_git_info=False)
|
||||||
|
ex.observers.append(FileStorageObserver("lab07/my_runs"))
|
||||||
|
|
||||||
|
|
||||||
|
@ex.config
|
||||||
|
def my_config():
|
||||||
|
train_size_param = 0.8
|
||||||
|
test_size_param = 0.2
|
||||||
|
epochs = 400
|
||||||
|
batch_size = 128
|
||||||
|
|
||||||
|
|
||||||
|
@ex.capture
|
||||||
|
def prepare_model(train_size_param, test_size_param, epochs, batch_size, _run):
|
||||||
|
_run.info["prepare_model_ts"] = str(datetime.now())
|
||||||
|
movies_data = pd.read_csv("train.csv", error_bad_lines=False)
|
||||||
|
movies_data.drop(movies_data.columns[0], axis=1, inplace=True)
|
||||||
|
movies_data.dropna(inplace=True)
|
||||||
|
X = movies_data.drop("rating", axis=1)
|
||||||
|
Y = movies_data["rating"]
|
||||||
|
|
||||||
|
print(X, Y.values)
|
||||||
|
# Split set to train/test 8:2 ratio
|
||||||
|
X_train, X_test, Y_train, Y_test = train_test_split(
|
||||||
|
X, Y, test_size=test_size_param, random_state=42
|
||||||
|
)
|
||||||
|
|
||||||
|
test_df = pd.read_csv("test.csv")
|
||||||
|
test_df.drop(test_df.columns[0], axis=1, inplace=True)
|
||||||
|
x_test = test_df.drop("rating", axis=1)
|
||||||
|
y_test = test_df["rating"]
|
||||||
|
|
||||||
|
# Set up model
|
||||||
|
model = Sequential()
|
||||||
|
model.add(Dense(8, activation="relu"))
|
||||||
|
model.add(Dropout(0.5))
|
||||||
|
model.add(Dense(3, activation="relu"))
|
||||||
|
model.add(Dropout(0.5))
|
||||||
|
model.add(Dense(1))
|
||||||
|
model.compile(optimizer="adam", loss="mse")
|
||||||
|
|
||||||
|
early_stop = EarlyStopping(monitor="val_loss", mode="min", verbose=1, patience=10)
|
||||||
|
|
||||||
|
model.fit(
|
||||||
|
x=X_train.values,
|
||||||
|
y=Y_train.values,
|
||||||
|
validation_data=(X_test, Y_test.values),
|
||||||
|
batch_size=batch_size,
|
||||||
|
epochs=epochs,
|
||||||
|
callbacks=[early_stop],
|
||||||
|
)
|
||||||
|
y_pred = model.predict(x_test.values)
|
||||||
|
|
||||||
|
rmse = mean_squared_error(y_test, y_pred)
|
||||||
|
|
||||||
|
_run.info["Final Results: "] = rmse
|
||||||
|
|
||||||
|
model.save("model_movies")
|
||||||
|
|
||||||
|
return rmse
|
||||||
|
|
||||||
|
|
||||||
|
@ex.automain
|
||||||
|
def my_main(train_size_param, test_size_param, epochs, batch_size):
|
||||||
|
print(prepare_model())
|
||||||
|
|
||||||
|
|
||||||
|
r = ex.run()
|
||||||
|
ex.add_artifact("model_movies/saved_model.pb")
|
@ -2,41 +2,20 @@ import sys
|
|||||||
import mlflow
|
import mlflow
|
||||||
import pandas as pd
|
import pandas as pd
|
||||||
|
|
||||||
from sacred import Experiment
|
|
||||||
from sacred.observers import FileStorageObserver
|
|
||||||
from sklearn.metrics import mean_squared_error
|
from sklearn.metrics import mean_squared_error
|
||||||
from sklearn.model_selection import train_test_split
|
from sklearn.model_selection import train_test_split
|
||||||
from tensorflow.keras.callbacks import EarlyStopping
|
from tensorflow.keras.callbacks import EarlyStopping
|
||||||
from tensorflow.keras.layers import Dense, Dropout
|
from tensorflow.keras.layers import Dense, Dropout
|
||||||
from tensorflow.keras.models import Sequential
|
from tensorflow.keras.models import Sequential
|
||||||
|
|
||||||
'''
|
|
||||||
If you want to use it as sacred, you have to
|
|
||||||
uncomment lines, for now it's set up for MLFlow
|
|
||||||
'''
|
|
||||||
# ex = Experiment("file_observer", interactive=False, save_git_info=False)
|
|
||||||
# ex.observers.append(FileStorageObserver('lab07/my_runs'))
|
|
||||||
|
|
||||||
|
def prepare_model(train_size_param, test_size_param, epochs, batch_size):
|
||||||
# @ex.config
|
|
||||||
# def my_config():
|
|
||||||
# train_size_param = 0.8
|
|
||||||
# test_size_param = 0.2
|
|
||||||
# epochs = 400
|
|
||||||
# batch_size = 128
|
|
||||||
|
|
||||||
|
|
||||||
# @ex.capture
|
|
||||||
def prepare_model(train_size_param, test_size_param, epochs, batch_size): # _run):
|
|
||||||
# _run.info["prepare_model_ts"] = str(datetime.now())
|
|
||||||
movies_data = pd.read_csv("train.csv", error_bad_lines=False)
|
movies_data = pd.read_csv("train.csv", error_bad_lines=False)
|
||||||
movies_data.drop(movies_data.columns[0], axis=1, inplace=True)
|
movies_data.drop(movies_data.columns[0], axis=1, inplace=True)
|
||||||
movies_data.dropna(inplace=True)
|
movies_data.dropna(inplace=True)
|
||||||
X = movies_data.drop("rating", axis=1)
|
X = movies_data.drop("rating", axis=1)
|
||||||
Y = movies_data["rating"]
|
Y = movies_data["rating"]
|
||||||
|
|
||||||
print(X, Y.values)
|
|
||||||
# Split set to train/test 8:2 ratio
|
|
||||||
X_train, X_test, Y_train, Y_test = train_test_split(
|
X_train, X_test, Y_train, Y_test = train_test_split(
|
||||||
X, Y, test_size=test_size_param, random_state=42
|
X, Y, test_size=test_size_param, random_state=42
|
||||||
)
|
)
|
||||||
@ -69,22 +48,11 @@ def prepare_model(train_size_param, test_size_param, epochs, batch_size): # _ru
|
|||||||
|
|
||||||
rmse = mean_squared_error(y_test, y_pred)
|
rmse = mean_squared_error(y_test, y_pred)
|
||||||
|
|
||||||
# _run.info["Final Results: "] = rmse
|
|
||||||
|
|
||||||
model.save("model_movies")
|
model.save("model_movies")
|
||||||
|
|
||||||
return model, rmse
|
return model, rmse
|
||||||
|
|
||||||
|
|
||||||
# @ex.automain
|
|
||||||
# def my_main(train_size_param, test_size_param, epochs, batch_size):
|
|
||||||
# print(prepare_model())
|
|
||||||
|
|
||||||
|
|
||||||
# r = ex.run()
|
|
||||||
# ex.add_artifact("model_movies/saved_model.pb")
|
|
||||||
|
|
||||||
|
|
||||||
train_size_param = float(sys.argv[1]) if len(sys.argv) > 1 else 0.8
|
train_size_param = float(sys.argv[1]) if len(sys.argv) > 1 else 0.8
|
||||||
test_size_param = float(sys.argv[2]) if len(sys.argv) > 1 else 0.2
|
test_size_param = float(sys.argv[2]) if len(sys.argv) > 1 else 0.2
|
||||||
epochs = int(sys.argv[3]) if len(sys.argv) > 1 else 400
|
epochs = int(sys.argv[3]) if len(sys.argv) > 1 else 400
|
Loading…
Reference in New Issue
Block a user