robocars/robocar-training
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

feat: add option to build linear model

Browse Source
This commit is contained in:
Cyrille Nofficial 2022-06-09 09:36:46 +02:00
parent ad3eb16333
commit 91758cfffa
1 changed files with 22 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
src/tf_container/train.py
View File

@ -19,6 +19,9 @@ from tensorflow.keras.models import Model
from tensorflow.keras.preprocessing.image import load_img, img_to_array from tensorflow.keras.preprocessing.image import load_img, img_to_array
from tensorflow.python.client import device_lib from tensorflow.python.client import device_lib
MODEL_CATEGORICAL = "categorical"
MODEL_LINEAR = "linear"
def linear_bin(a: float, N: int = 15, offset: int = 1, R: float = 2.0): def linear_bin(a: float, N: int = 15, offset: int = 1, R: float = 2.0):
""" """
@ -57,7 +60,7 @@ def unzip_file(root, f):
zip_ref.close() zip_ref.close()
def train(batch_size: int, slide_size: int, img_height: int, img_width: int, img_depth: int, horizon: int, drop: float): def train(model_type: str, batch_size: int, slide_size: int, img_height: int, img_width: int, img_depth: int, horizon: int, drop: float):
# env = cs.TrainingEnvironment() # env = cs.TrainingEnvironment()
print(device_lib.list_local_devices()) print(device_lib.list_local_devices())
@ -108,7 +111,13 @@ def train(batch_size: int, slide_size: int, img_height: int, img_width: int, img
# imgs = np.reshape(images[0:25], (-1, img_height, img_width, img_depth)) # imgs = np.reshape(images[0:25], (-1, img_height, img_width, img_depth))
# tf.summary.image("25 training data examples", imgs, max_outputs=25, step=0) # tf.summary.image("25 training data examples", imgs, max_outputs=25, step=0)
save_best = callbacks.ModelCheckpoint('/opt/ml/model/model_cat', monitor='val_loss', verbose=1, model_filepath = '/opt/ml/model/model_other'
if model_type == MODEL_CATEGORICAL:
model_filepath = '/opt/ml/model/model_cat'
elif model_type == MODEL_LINEAR:
model_filepath = '/opt/ml/model/model_lin'
save_best = callbacks.ModelCheckpoint(model_filepath, monitor='val_loss', verbose=1,
save_best_only=True, mode='min') save_best_only=True, mode='min')
early_stop = callbacks.EarlyStopping(monitor='val_loss', early_stop = callbacks.EarlyStopping(monitor='val_loss',
min_delta=.0005, min_delta=.0005,
@ -121,8 +130,8 @@ def train(batch_size: int, slide_size: int, img_height: int, img_width: int, img
angle_cat_array = np.array([linear_bin(float(a)) for a in angle_array]) angle_cat_array = np.array([linear_bin(float(a)) for a in angle_array])
model = default_model(input_shape=(img_height - horizon, img_width, img_depth), drop=drop) #model = default_model(input_shape=(img_height - horizon, img_width, img_depth), drop=drop)
#model = default_categorical(input_shape=(img_height - horizon, img_width, img_depth), drop=drop) model = default_categorical(input_shape=(img_height - horizon, img_width, img_depth), drop=drop)
model.compile(optimizer='adam', model.compile(optimizer='adam',
loss={'angle_out': 'categorical_crossentropy', }, loss={'angle_out': 'categorical_crossentropy', },
@ -150,7 +159,7 @@ def train(batch_size: int, slide_size: int, img_height: int, img_width: int, img
tflite_model = converter.convert() tflite_model = converter.convert()
# Save the model. # Save the model.
with open('/opt/ml/model/model_' + str(img_width) + 'x' + str(img_height) + 'h' + str(horizon) + '.tflite', with open('/opt/ml/model/model_' + model_type + '_' + str(img_width) + 'x' + str(img_height) + 'h' + str(horizon) + '.tflite',
'wb') as f: 'wb') as f:
f.write(tflite_model) f.write(tflite_model)
@ -177,6 +186,8 @@ def core_cnn_layers(img_in: Input, img_height: int, img_width: int, drop: float,
""" """
Returns the core CNN layers that are shared among the different models, Returns the core CNN layers that are shared among the different models,
like linear, imu, behavioural like linear, imu, behavioural
:param img_width: image width
:param img_height: image height
:param img_in: input layer of network :param img_in: input layer of network
:param drop: dropout rate :param drop: dropout rate
:param l4_stride: 4-th layer stride, default 1 :param l4_stride: 4-th layer stride, default 1
@ -236,20 +247,16 @@ def default_model(input_shape, drop):
return model return model
def default_n_linear(num_outputs, input_shape=(120, 160, 3), drop=0.2): def default_linear(input_shape=(120, 160, 3), drop=0.2):
img_in = Input(shape=input_shape, name='img_in') img_in = Input(shape=input_shape, name='img_in')
x = core_cnn_layers(img_in, img_width=input_shape[1], img_height=input_shape[0], drop=drop) x = core_cnn_layers(img_in, img_width=input_shape[1], img_height=input_shape[0], drop=drop)
x = Dense(100, activation='relu', name='dense_1')(x) x = Dense(100, activation='relu', name='dense_1')(x)
x = Dropout(drop)(x) x = Dropout(drop)(x)
x = Dense(50, activation='relu', name='dense_2')(x) x = Dense(50, activation='relu', name='dense_2')(x)
x = Dropout(drop)(x) x = Dropout(drop)(x)
angle_out = Dense(1, activation='linear', name='angle_out')(x)
outputs = [] model = Model(inputs=[img_in], outputs=[angle_out], name='linear')
for i in range(num_outputs):
outputs.append(
Dense(1, activation='linear', name='n_outputs' + str(i))(x))
model = Model(inputs=[img_in], outputs=outputs, name='linear')
return model return model
@ -263,8 +270,7 @@ def default_categorical(input_shape=(120, 160, 3), drop=0.2):
# Categorical output of the angle into 15 bins # Categorical output of the angle into 15 bins
angle_out = Dense(15, activation='softmax', name='angle_out')(x) angle_out = Dense(15, activation='softmax', name='angle_out')(x)
model = Model(inputs=[img_in], outputs=[angle_out], model = Model(inputs=[img_in], outputs=[angle_out], name='categorical')
name='categorical')
return model return model
@ -278,10 +284,12 @@ if __name__ == "__main__":
parser.add_argument("--horizon", type=int, default=0) parser.add_argument("--horizon", type=int, default=0)
parser.add_argument("--batch_size", type=int, default=32) parser.add_argument("--batch_size", type=int, default=32)
parser.add_argument("--drop", type=float, default=0.2) parser.add_argument("--drop", type=float, default=0.2)
parser.add_argument("--model_type", type=str, default=MODEL_CATEGORICAL)
args = parser.parse_args() args = parser.parse_args()
params = vars(args) params = vars(args)
train( train(
model_type=params["model_type"],
batch_size=params["batch_size"], batch_size=params["batch_size"],
slide_size=params["slide_size"], slide_size=params["slide_size"],
img_height=params["img_height"], img_height=params["img_height"],