search

bendangnuksung / mrcnn_serving_ready

đŸ›  Mask R-CNN Keras to Tensorflow and TFX models + Serving models using TFX GRPC & RESTAPI
https://github.com/matterport/Mask_RCNN
MIT License
103 stars 35 forks source link

I try to run with mask_rcnn_coco.h5 #1

Closed NorthLatitudeOne closed 5 years ago

NorthLatitudeOne commented 5 years ago

Hi , bendangnuksung, I try to run with mask_rcnn_coco.h5, but I got the error message as below: name: GeForce GTX 1080 Ti major: 6 minor: 1 memoryClockRate(GHz): 1.721 pciBusID: 0000:01:00.0 totalMemory: 11.00GiB freeMemory: 9.10GiB 2019-04-15 08:51:08.395483: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1511] Adding visible gpu devices: 0 2019-04-15 08:51:09.217262: I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] Device interconnect StreamExecutor with strength 1 edge matrix: 2019-04-15 08:51:09.222426: I tensorflow/core/common_runtime/gpu/gpu_device.cc:988] 0 2019-04-15 08:51:09.225392: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1001] 0: N 2019-04-15 08:51:09.228476: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 8788 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1080 Ti, pci bus id: 0000:01:00.0, compute capability: 6.1) WARNING:tensorflow:From E:\Anaconda\lib\site-packages\tensorflow\python\ops\sparse_ops.py:1165: sparse_to_dense (from tensorflow.python.ops.sparse_ops) is deprecated and will be removed in a future version. Instructions for updating: Create atf.sparse.SparseTensorand usetf.sparse.to_denseinstead. Traceback (most recent call last): File "main.py", line 88, in <module> model.load_weights(H5_WEIGHT_PATH, by_name=True) File "F:\Code\mrcnn_serving_ready-master\model.py", line 2131, in load_weights saving.load_weights_from_hdf5_group_by_name(f, layers) File "E:\Anaconda\lib\site-packages\keras\engine\saving.py", line 1149, in load_weights_from_hdf5_group_by_name str(weight_values[i].shape) + '.') ValueError: Layer #389 (named "mrcnn_bbox_fc"), weight <tf.Variable 'mrcnn_bbox_fc/kernel:0' shape=(1024, 24) dtype=float32_ref> has shape (1024, 24), but the saved weight has shape (1024, 324).

bendangnuksung commented 5 years ago

@NorthLatitudeOne It was happening because COCO model uses a little different config. Made an update, you can convert 'mask_rcnn_coco.h5'. You just need open 'user_config.py' and set

# Make it True if you want to use the provided coco weights
is_coco = True
NorthLatitudeOne commented 5 years ago

Hi Bendang , thanks for your help, the problem solved now after adding the config.

buaacarzp commented 5 years ago

Hi Bendang , thanks for your help, the problem solved now after adding the config.

Hi,are you ok now? did you completed the mrcnn_coco.h5 to pb file ? why I always failure.... please help me .