multi gpu training error

[iRedRum]

Hi, thanks for the sharing of the great work.

I am kind of stuck in multigpu training with the model. I got the error: Traceback (most recent call last): File "coco.py", line 567, in <module> model_dir=args.logs) File "/home/yanyy/yan/Keypoints-of-humanpose-with-Mask-R-CNN/model.py", line 2538, in __init__ self.keras_model = self.build(mode=mode, config=config) File "/home/yanyy/yan/Keypoints-of-humanpose-with-Mask-R-CNN/model.py", line 2811, in build model = ParallelModel(model, config.GPU_COUNT) File "/home/yanyy/yan/Keypoints-of-humanpose-with-Mask-R-CNN/parallel_model.py", line 37, in __init__ merged_outputs = self.make_parallel() File "/home/yanyy/yan/Keypoints-of-humanpose-with-Mask-R-CNN/parallel_model.py", line 81, in make_parallel outputs = self.inner_model(inputs) File "/home/yanyy/anaconda3/envs/tf/lib/python3.5/site-packages/keras/engine/topology.py", line 603, in __call__ output = self.call(inputs, **kwargs) File "/home/yanyy/anaconda3/envs/tf/lib/python3.5/site-packages/keras/engine/topology.py", line 2061, in call output_tensors, _, _ = self.run_internal_graph(inputs, masks) File "/home/yanyy/anaconda3/envs/tf/lib/python3.5/site-packages/keras/engine/topology.py", line 2265, in run_internal_graph assert str(id(x)) in tensor_map, 'Could not compute output ' + str(x) AssertionError: Could not compute output Tensor("mrcnn_class_loss/Reshape:0", shape=(1, 1), dtype=float32)

Did you try to train the model on the multi-gpu

[HONGJUNRL]

Did you solve this problem?

[hci-mkim]

how did you solve this problem?