qurator-spk/eynollah
1
0
Fork 0
mirror of https://github.com/qurator-spk/eynollah.git synced 2026-02-20 16:32:03 +01:00
Code Issues Projects Releases Wiki Activity

training.train: fix F1 metric score setup

Browse source
This commit is contained in:
Robert Sachunsky 2026-02-05 11:57:38 +01:00
parent 5c7801a1d6
commit 82d649061a
1 changed files with 9 additions and 8 deletions
Download patch file Download diff file Expand all files Collapse all files

17
src/eynollah/training/train.py
View file

@ -34,7 +34,7 @@ os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
os.environ['TF_USE_LEGACY_KERAS'] = '1' # avoid Keras 3 after TF 2.15 os.environ['TF_USE_LEGACY_KERAS'] = '1' # avoid Keras 3 after TF 2.15
import tensorflow as tf import tensorflow as tf
from tensorflow.keras.optimizers import SGD, Adam from tensorflow.keras.optimizers import SGD, Adam
from tensorflow.keras.metrics import MeanIoU from tensorflow.keras.metrics import MeanIoU, F1Score
from tensorflow.keras.models import load_model from tensorflow.keras.models import load_model
from tensorflow.keras.callbacks import ModelCheckpoint, TensorBoard from tensorflow.keras.callbacks import ModelCheckpoint, TensorBoard
from sacred import Experiment from sacred import Experiment
@ -427,8 +427,8 @@ def run(_config,
model.compile(loss='categorical_crossentropy', model.compile(loss='categorical_crossentropy',
optimizer=Adam(learning_rate=0.001), # rs: why not learning_rate? optimizer=Adam(learning_rate=0.001), # rs: why not learning_rate?
metrics=['accuracy']) metrics=['accuracy', F1Score(average='macro', name='f1')])
list_classes = list(classification_classes_name.values()) list_classes = list(classification_classes_name.values())
trainXY = generate_data_from_folder_training( trainXY = generate_data_from_folder_training(
dir_train, n_batch, input_height, input_width, n_classes, list_classes) dir_train, n_batch, input_height, input_width, n_classes, list_classes)
@ -440,7 +440,8 @@ def run(_config,
callbacks = [TensorBoard(os.path.join(dir_output, 'logs'), write_graph=False), callbacks = [TensorBoard(os.path.join(dir_output, 'logs'), write_graph=False),
SaveWeightsAfterSteps(0, dir_output, _config, SaveWeightsAfterSteps(0, dir_output, _config,
monitor='val_f1', monitor='val_f1',
save_best_only=True, mode='max')] #save_best_only=True, # we need all for ensembling
mode='max')]
history = model.fit(trainXY, history = model.fit(trainXY,
steps_per_epoch=num_rows / n_batch, steps_per_epoch=num_rows / n_batch,
@ -448,17 +449,17 @@ def run(_config,
validation_data=testXY, validation_data=testXY,
verbose=1, verbose=1,
epochs=n_epochs, epochs=n_epochs,
metrics=[F1Score(average='macro', name='f1')],
callbacks=callbacks, callbacks=callbacks,
initial_epoch=index_start) initial_epoch=index_start)
usable_checkpoints = np.flatnonzero(np.array(history['val_f1']) > f1_threshold_classification) usable_checkpoints = np.flatnonzero(np.array(history.history['val_f1']) >
f1_threshold_classification)
if len(usable_checkpoints) >= 1: if len(usable_checkpoints) >= 1:
_log.info("averaging over usable checkpoints: %s", str(usable_checkpoints)) _log.info("averaging over usable checkpoints: %s", str(usable_checkpoints))
all_weights = [] all_weights = []
for epoch in usable_checkpoints: for epoch in usable_checkpoints:
cp_path = os.path.join(dir_output, 'model_{epoch:02d}'.format(epoch=epoch)) cp_path = os.path.join(dir_output, 'model_{epoch:02d}'.format(epoch=epoch + 1))
assert os.path.isdir(cp_path) assert os.path.isdir(cp_path), cp_path
model = load_model(cp_path, compile=False) model = load_model(cp_path, compile=False)
all_weights.append(model.get_weights()) all_weights.append(model.get_weights())