apple / tensorflow_macos

TensorFlow for macOS 11.0+ accelerated using Apple's ML Compute framework.
Other
3.66k stars 308 forks source link

Unable to Run Object Detection Training #171

Open bks715 opened 3 years ago

bks715 commented 3 years ago

When I try to run object detection training, I get this error. I'm using Python 3.8.7 on a 16" MacBook Pro running macOS 11.2. I'm assuming it's because there isn't a matching layer for this "ChangeCoordinateFrame/Scale/mul (Mul)" but I'm not entirely sure.

2021-02-15 15:42:57.097358: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:116] None of the MLIR optimization passes are enabled (registered 2) Fatal Python error: Segmentation fault

Thread 0x000000010e965e00 (most recent call first): File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/ops/gen_dataset_ops.py", line 2574 in iterator_get_next File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/data/ops/iterator_ops.py", line 730 in _next_internal File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/data/ops/iterator_ops.py", line 800 in get_next File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/data/ops/multi_device_iterator_ops.py", line 585 in get_2021-02-15 15:42:57.525030: I tensorflow/compiler/tf2mlcompute/kernels/mlc_subgraph_op.cc:332] Compute: Failed in processing TensorFlow graph MLCSubgraphOp_0_0 with frame_id = 0 and iter_id = 0 with error: Internal: AddLayerToMLCGraph: Failed to add layer to MLCGraph for node: n19 = ChangeCoordinateFrame/Scale/mul (Mul). (error will be reported 5 times unless TF_MLC_LOGGING=1). next File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/distribute/input_lib.py", line 1610 in get_next_as_list_static_shapes File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/distribute/input_lib.py", line 662 in get_next File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/distribute/input_lib.py", line 631 in next File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/distribute/input_lib.py", line 627 in next File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/object_detection/model_lib_v2.py", line 352 in load_fine_tune_checkpoint File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/object_detection/model_lib_v2.py", line 569 in train_loop File "object_detection/model_main_tf2.py", line 117 in main File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/absl/app.py", line 251 in _run_main File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/absl/app.py", line 300 in run File "/Users/bradensmith/tensorflow_macos_venv/lib/python3.8/site-packages/tensorflow/python/platform/app.py", line 40 in run File "object_detection/model_main_tf2.py", line 126 in zsh: segmentation fault python object_detection/model_main_tf2.py --alsologtostderr --use_tpu=false

anna-tikhonova commented 3 years ago

Thank you very much for reporting this issue. Could you please provide a reproducible test case, so we can run it locally and investigate?