TypeError: reduce_mean() got an unexpected keyword argument 'keepdims'

[ZengyuanYu]

Like this , I am training bollon in GPUs occer this issues,in detiles: Training network heads

Starting at epoch 0. LR=0.001

Checkpoint Path: /home/yu/Mask_RCNN/logs/balloon20180517T2146/mask_rcnnballoon{epoch:04d}.h5 Selecting layers to train fpn_c5p5 (Conv2D) fpn_c4p4 (Conv2D) fpn_c3p3 (Conv2D) fpn_c2p2 (Conv2D) fpn_p5 (Conv2D) fpn_p2 (Conv2D) fpn_p3 (Conv2D) fpn_p4 (Conv2D) In model: rpn_model rpn_conv_shared (Conv2D) rpn_class_raw (Conv2D) rpn_bbox_pred (Conv2D) mrcnn_mask_conv1 (TimeDistributed) mrcnn_mask_bn1 (TimeDistributed) mrcnn_mask_conv2 (TimeDistributed) mrcnn_mask_bn2 (TimeDistributed) mrcnn_class_conv1 (TimeDistributed) mrcnn_class_bn1 (TimeDistributed) mrcnn_mask_conv3 (TimeDistributed) mrcnn_mask_bn3 (TimeDistributed) mrcnn_class_conv2 (TimeDistributed) mrcnn_class_bn2 (TimeDistributed) mrcnn_mask_conv4 (TimeDistributed) mrcnn_mask_bn4 (TimeDistributed) mrcnn_bbox_fc (TimeDistributed) mrcnn_mask_deconv (TimeDistributed) mrcnn_class_logits (TimeDistributed) mrcnn_mask (TimeDistributed) Traceback (most recent call last): File "balloon.py", line 359, in train(model) File "balloon.py", line 194, in train layers='heads') File "/home/yu/.virtualenvs/Mask_RCNN/local/lib/python3.5/site-packages/mask_rcnn-2.1-py3.5.egg/mrcnn/model.py", line 2308, in train File "/home/yu/.virtualenvs/Mask_RCNN/local/lib/python3.5/site-packages/mask_rcnn-2.1-py3.5.egg/mrcnn/model.py", line 2143, in compile TypeError: reduce_mean() got an unexpected keyword argument 'keepdims'

tf = 1.4 keras =2.0.8 I hope anyone fix it .thank you!

[patrick-llgc]

See this commit. You can go to mrcnn/model.py and change it back to keep_dims or update to tf 1.5+.

[celsomilne]

Should requirements.txt be updated to reflect the need for tf>=1.5?