tensorflow.python.framework.errors_impl.InvalidArgumentError: Invalid argument: assertion failed:

Question

While training using TensorFlow Object Detection API from Google Colab I got the following error (There are two similar errors in the following verbose..one of them is in the end of it):

WARNING:tensorflow:Forced number of epochs for all eval validations to be 1.
W0528 21:13:21.113062 140292083513216 model_lib.py:717] Forced number of epochs for all eval validations to be 1.
INFO:tensorflow:Maybe overwriting train_steps: 200000
I0528 21:13:21.113316 140292083513216 config_util.py:523] Maybe overwriting train_steps: 200000
INFO:tensorflow:Maybe overwriting use_bfloat16: False
I0528 21:13:21.113430 140292083513216 config_util.py:523] Maybe overwriting use_bfloat16: False
INFO:tensorflow:Maybe overwriting sample_1_of_n_eval_examples: 1
I0528 21:13:21.113519 140292083513216 config_util.py:523] Maybe overwriting sample_1_of_n_eval_examples: 1
INFO:tensorflow:Maybe overwriting eval_num_epochs: 1
I0528 21:13:21.113614 140292083513216 config_util.py:523] Maybe overwriting eval_num_epochs: 1
INFO:tensorflow:Maybe overwriting load_pretrained: True
I0528 21:13:21.113696 140292083513216 config_util.py:523] Maybe overwriting load_pretrained: True
INFO:tensorflow:Ignoring config override key: load_pretrained
I0528 21:13:21.113776 140292083513216 config_util.py:533] Ignoring config override key: load_pretrained
WARNING:tensorflow:Expected number of evaluation epochs is 1, but instead encountered `eval_on_train_input_config.num_epochs` = 0. Overwriting `num_epochs` to 1.
W0528 21:13:21.114626 140292083513216 model_lib.py:733] Expected number of evaluation epochs is 1, but instead encountered `eval_on_train_input_config.num_epochs` = 0. Overwriting `num_epochs` to 1.
INFO:tensorflow:create_estimator_and_inputs: use_tpu False, export_to_tpu False
I0528 21:13:21.114744 140292083513216 model_lib.py:768] create_estimator_and_inputs: use_tpu False, export_to_tpu False
INFO:tensorflow:Using config: {'_model_dir': 'training/', '_tf_random_seed': None, '_save_summary_steps': 100, '_save_checkpoints_steps': None, '_save_checkpoints_secs': 600, '_session_config': allow_soft_placement: true
graph_options {
  rewrite_options {
    meta_optimizer_iterations: ONE
  }
}
, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': 100, '_train_distribute': None, '_device_fn': None, '_protocol': None, '_eval_distribute': None, '_experimental_distribute': None, '_experimental_max_worker_delay_secs': None, '_session_creation_timeout_secs': 7200, '_service': None, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f97ed4dd128>, '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1}
I0528 21:13:21.115245 140292083513216 estimator.py:212] Using config: {'_model_dir': 'training/', '_tf_random_seed': None, '_save_summary_steps': 100, '_save_checkpoints_steps': None, '_save_checkpoints_secs': 600, '_session_config': allow_soft_placement: true
graph_options {
  rewrite_options {
    meta_optimizer_iterations: ONE
  }
}
, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': 100, '_train_distribute': None, '_device_fn': None, '_protocol': None, '_eval_distribute': None, '_experimental_distribute': None, '_experimental_max_worker_delay_secs': None, '_session_creation_timeout_secs': 7200, '_service': None, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f97ed4dd128>, '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1}
WARNING:tensorflow:Estimator's model_fn (<function create_model_fn.<locals>.model_fn at 0x7f97d328dbf8>) includes params argument, but params are not passed to Estimator.
W0528 21:13:21.115487 140292083513216 model_fn.py:630] Estimator's model_fn (<function create_model_fn.<locals>.model_fn at 0x7f97d328dbf8>) includes params argument, but params are not passed to Estimator.
INFO:tensorflow:Not using Distribute Coordinator.
I0528 21:13:21.116259 140292083513216 estimator_training.py:186] Not using Distribute Coordinator.
INFO:tensorflow:Running training and evaluation locally (non-distributed).
I0528 21:13:21.116456 140292083513216 training.py:612] Running training and evaluation locally (non-distributed).
INFO:tensorflow:Start train and evaluate loop. The evaluate will happen after every checkpoint. Checkpoint frequency is determined based on RunConfig arguments: save_checkpoints_steps None or save_checkpoints_secs 600.
I0528 21:13:21.116694 140292083513216 training.py:700] Start train and evaluate loop. The evaluate will happen after every checkpoint. Checkpoint frequency is determined based on RunConfig arguments: save_checkpoints_steps None or save_checkpoints_secs 600.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/training_util.py:236: Variable.initialized_value (from tensorflow.python.ops.variables) is deprecated and will be removed in a future version.
Instructions for updating:
Use Variable.read_value. Variables in 2.X are initialized automatically both in eager and graph (inside tf.defun) contexts.
W0528 21:13:21.124795 140292083513216 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/training_util.py:236: Variable.initialized_value (from tensorflow.python.ops.variables) is deprecated and will be removed in a future version.
Instructions for updating:
Use Variable.read_value. Variables in 2.X are initialized automatically both in eager and graph (inside tf.defun) contexts.
WARNING:tensorflow:num_readers has been reduced to 1 to match input file shards.
W0528 21:13:21.162153 140292083513216 dataset_builder.py:84] num_readers has been reduced to 1 to match input file shards.
WARNING:tensorflow:From /content/models/research/object_detection/builders/dataset_builder.py:101: parallel_interleave (from tensorflow.contrib.data.python.ops.interleave_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.data.experimental.parallel_interleave(...)`.
W0528 21:13:21.167545 140292083513216 deprecation.py:323] From /content/models/research/object_detection/builders/dataset_builder.py:101: parallel_interleave (from tensorflow.contrib.data.python.ops.interleave_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.data.experimental.parallel_interleave(...)`.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/contrib/data/python/ops/interleave_ops.py:77: parallel_interleave (from tensorflow.python.data.experimental.ops.interleave_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.data.Dataset.interleave(map_func, cycle_length, block_length, num_parallel_calls=tf.data.experimental.AUTOTUNE)` instead. If sloppy execution is desired, use `tf.data.Options.experimental_determinstic`.
W0528 21:13:21.167754 140292083513216 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow_core/contrib/data/python/ops/interleave_ops.py:77: parallel_interleave (from tensorflow.python.data.experimental.ops.interleave_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.data.Dataset.interleave(map_func, cycle_length, block_length, num_parallel_calls=tf.data.experimental.AUTOTUNE)` instead. If sloppy execution is desired, use `tf.data.Options.experimental_determinstic`.
2020-05-28 21:13:22.910301: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcuda.so.1
2020-05-28 21:13:22.953259: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:13:22.953875: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1618] Found device 0 with properties: 
name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59
pciBusID: 0000:00:04.0
2020-05-28 21:13:22.960996: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.0
2020-05-28 21:13:22.967688: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10.0
2020-05-28 21:13:22.977811: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10.0
2020-05-28 21:13:22.985131: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10.0
2020-05-28 21:13:22.995549: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10.0
2020-05-28 21:13:23.004617: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10.0
2020-05-28 21:13:23.025234: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7
2020-05-28 21:13:23.025382: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:13:23.026101: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:13:23.026693: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1746] Adding visible gpu devices: 0
WARNING:tensorflow:From /content/models/research/object_detection/inputs.py:77: sparse_to_dense (from tensorflow.python.ops.sparse_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Create a `tf.sparse.SparseTensor` and use `tf.sparse.to_dense` instead.
W0528 21:13:33.109247 140292083513216 deprecation.py:323] From /content/models/research/object_detection/inputs.py:77: sparse_to_dense (from tensorflow.python.ops.sparse_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Create a `tf.sparse.SparseTensor` and use `tf.sparse.to_dense` instead.
WARNING:tensorflow:From /content/models/research/object_detection/utils/ops.py:493: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.where in 2.0, which has the same broadcast rule as np.where
W0528 21:13:33.221111 140292083513216 deprecation.py:323] From /content/models/research/object_detection/utils/ops.py:493: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.where in 2.0, which has the same broadcast rule as np.where
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/autograph/operators/control_flow.py:1004: sample_distorted_bounding_box (from tensorflow.python.ops.image_ops_impl) is deprecated and will be removed in a future version.
Instructions for updating:
`seed2` arg is deprecated.Use sample_distorted_bounding_box_v2 instead.
W0528 21:13:39.145547 140292083513216 api.py:332] From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/autograph/operators/control_flow.py:1004: sample_distorted_bounding_box (from tensorflow.python.ops.image_ops_impl) is deprecated and will be removed in a future version.
Instructions for updating:
`seed2` arg is deprecated.Use sample_distorted_bounding_box_v2 instead.
WARNING:tensorflow:From /content/models/research/object_detection/inputs.py:259: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.cast` instead.
W0528 21:13:42.865469 140292083513216 deprecation.py:323] From /content/models/research/object_detection/inputs.py:259: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.cast` instead.
WARNING:tensorflow:From /content/models/research/object_detection/builders/dataset_builder.py:174: batch_and_drop_remainder (from tensorflow.contrib.data.python.ops.batching) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.data.Dataset.batch(..., drop_remainder=True)`.
W0528 21:13:46.217640 140292083513216 deprecation.py:323] From /content/models/research/object_detection/builders/dataset_builder.py:174: batch_and_drop_remainder (from tensorflow.contrib.data.python.ops.batching) is deprecated and will be removed in a future version.
Instructions for updating:
Use `tf.data.Dataset.batch(..., drop_remainder=True)`.
INFO:tensorflow:Calling model_fn.
I0528 21:13:46.233859 140292083513216 estimator.py:1148] Calling model_fn.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tf_slim/layers/layers.py:1089: Layer.apply (from tensorflow.python.keras.engine.base_layer) is deprecated and will be removed in a future version.
Instructions for updating:
Please use `layer.__call__` method instead.
W0528 21:13:46.430602 140292083513216 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tf_slim/layers/layers.py:1089: Layer.apply (from tensorflow.python.keras.engine.base_layer) is deprecated and will be removed in a future version.
Instructions for updating:
Please use `layer.__call__` method instead.
INFO:tensorflow:depth of additional conv before box predictor: 0
I0528 21:13:49.101978 140292083513216 convolutional_box_predictor.py:156] depth of additional conv before box predictor: 0
INFO:tensorflow:depth of additional conv before box predictor: 0
I0528 21:13:49.133970 140292083513216 convolutional_box_predictor.py:156] depth of additional conv before box predictor: 0
INFO:tensorflow:depth of additional conv before box predictor: 0
I0528 21:13:49.165436 140292083513216 convolutional_box_predictor.py:156] depth of additional conv before box predictor: 0
INFO:tensorflow:depth of additional conv before box predictor: 0
I0528 21:13:49.343221 140292083513216 convolutional_box_predictor.py:156] depth of additional conv before box predictor: 0
INFO:tensorflow:depth of additional conv before box predictor: 0
I0528 21:13:49.377842 140292083513216 convolutional_box_predictor.py:156] depth of additional conv before box predictor: 0
INFO:tensorflow:depth of additional conv before box predictor: 0
I0528 21:13:49.414346 140292083513216 convolutional_box_predictor.py:156] depth of additional conv before box predictor: 0
W0528 21:13:49.456603 140292083513216 variables_helper.py:161] Variable [FeatureExtractor/MobilenetV2/layer_19_2_Conv2d_2_3x3_s2_512/weights] is available in checkpoint, but has an incompatible shape with model variable. Checkpoint shape: [[1, 1, 256, 512]], model variable shape: [[3, 3, 256, 512]]. This variable will not be initialized from the checkpoint.
W0528 21:13:49.456816 140292083513216 variables_helper.py:161] Variable [FeatureExtractor/MobilenetV2/layer_19_2_Conv2d_3_3x3_s2_256/weights] is available in checkpoint, but has an incompatible shape with model variable. Checkpoint shape: [[1, 1, 128, 256]], model variable shape: [[3, 3, 128, 256]]. This variable will not be initialized from the checkpoint.
W0528 21:13:49.456997 140292083513216 variables_helper.py:161] Variable [FeatureExtractor/MobilenetV2/layer_19_2_Conv2d_4_3x3_s2_256/weights] is available in checkpoint, but has an incompatible shape with model variable. Checkpoint shape: [[1, 1, 128, 256]], model variable shape: [[3, 3, 128, 256]]. This variable will not be initialized from the checkpoint.
W0528 21:13:49.457174 140292083513216 variables_helper.py:161] Variable [FeatureExtractor/MobilenetV2/layer_19_2_Conv2d_5_3x3_s2_128/weights] is available in checkpoint, but has an incompatible shape with model variable. Checkpoint shape: [[1, 1, 64, 128]], model variable shape: [[3, 3, 64, 128]]. This variable will not be initialized from the checkpoint.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/rmsprop.py:119: calling Ones.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.
Instructions for updating:
Call initializer instance with the dtype argument instead of passing it to the constructor
W0528 21:13:54.449208 140292083513216 deprecation.py:506] From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/rmsprop.py:119: calling Ones.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.
Instructions for updating:
Call initializer instance with the dtype argument instead of passing it to the constructor
INFO:tensorflow:Done calling model_fn.
I0528 21:14:00.871218 140292083513216 estimator.py:1150] Done calling model_fn.
INFO:tensorflow:Create CheckpointSaverHook.
I0528 21:14:00.872715 140292083513216 basic_session_run_hooks.py:541] Create CheckpointSaverHook.
INFO:tensorflow:Graph was finalized.
I0528 21:14:04.557027 140292083513216 monitored_session.py:240] Graph was finalized.
2020-05-28 21:14:04.557485: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 AVX512F FMA
2020-05-28 21:14:04.562729: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2000165000 Hz
2020-05-28 21:14:04.563012: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x1771800 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-05-28 21:14:04.563048: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version
2020-05-28 21:14:04.666903: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:14:04.667672: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x1770d80 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2020-05-28 21:14:04.667705: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Tesla T4, Compute Capability 7.5
2020-05-28 21:14:04.668018: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:14:04.668594: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1618] Found device 0 with properties: 
name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59
pciBusID: 0000:00:04.0
2020-05-28 21:14:04.668682: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.0
2020-05-28 21:14:04.668724: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10.0
2020-05-28 21:14:04.668747: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10.0
2020-05-28 21:14:04.668769: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10.0
2020-05-28 21:14:04.668796: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10.0
2020-05-28 21:14:04.668819: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10.0
2020-05-28 21:14:04.668842: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7
2020-05-28 21:14:04.668951: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:14:04.669555: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:14:04.670109: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1746] Adding visible gpu devices: 0
2020-05-28 21:14:04.670229: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.0
2020-05-28 21:14:04.671546: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1159] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-05-28 21:14:04.671575: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1165]      0 
2020-05-28 21:14:04.671585: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1178] 0:   N 
2020-05-28 21:14:04.671747: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:14:04.672416: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:983] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-05-28 21:14:04.672994: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:39] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.
2020-05-28 21:14:04.673037: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1304] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14221 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)
INFO:tensorflow:Running local_init_op.
I0528 21:14:09.605103 140292083513216 session_manager.py:500] Running local_init_op.
INFO:tensorflow:Done running local_init_op.
I0528 21:14:09.941666 140292083513216 session_manager.py:502] Done running local_init_op.
INFO:tensorflow:Saving checkpoints for 0 into training/model.ckpt.
I0528 21:14:18.960145 140292083513216 basic_session_run_hooks.py:606] Saving checkpoints for 0 into training/model.ckpt.
2020-05-28 21:14:36.916392: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:145] Filling up shuffle buffer (this may take a while): 1074 of 2048
2020-05-28 21:14:46.905139: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:145] Filling up shuffle buffer (this may take a while): 2026 of 2048
2020-05-28 21:14:46.910085: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:195] Shuffle buffer filled.
2020-05-28 21:14:47.284742: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7
2020-05-28 21:14:53.420068: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10.0
INFO:tensorflow:loss = 12.133639, step = 0
I0528 21:14:56.692664 140292083513216 basic_session_run_hooks.py:262] loss = 12.133639, step = 0
Traceback (most recent call last):
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/client/session.py", line 1365, in _do_call
    return fn(*args)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/client/session.py", line 1350, in _run_fn
    target_list, run_metadata)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/client/session.py", line 1443, in _call_tf_sessionrun
    run_metadata)
tensorflow.python.framework.errors_impl.InvalidArgumentError: 2 root error(s) found.
  (0) Invalid argument: {{function_node __inference_Dataset_map_transform_and_pad_input_data_fn_3047}} assertion failed: [[0.748][0.758]] [[0.67][0.67]]
     [[{{node Assert/AssertGuard/else/_123/Assert}}]]
     [[IteratorGetNext]]
  (1) Invalid argument: {{function_node __inference_Dataset_map_transform_and_pad_input_data_fn_3047}} assertion failed: [[0.748][0.758]] [[0.67][0.67]]
     [[{{node Assert/AssertGuard/else/_123/Assert}}]]
     [[IteratorGetNext]]
     [[IteratorGetNext/_8451]]
0 successful operations.
0 derived errors ignored.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/content/models/research/object_detection/model_main.py", line 114, in <module>
    tf.app.run()
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/platform/app.py", line 40, in run
    _run(main=main, argv=argv, flags_parser=_parse_flags_tolerate_undef)
  File "/usr/local/lib/python3.6/dist-packages/absl/app.py", line 299, in run
    _run_main(main, args)
  File "/usr/local/lib/python3.6/dist-packages/absl/app.py", line 250, in _run_main
    sys.exit(main(argv))
  File "/content/models/research/object_detection/model_main.py", line 110, in main
    tf.estimator.train_and_evaluate(estimator, train_spec, eval_specs[0])
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_estimator/python/estimator/training.py", line 473, in train_and_evaluate
    return executor.run()
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_estimator/python/estimator/training.py", line 613, in run
    return self.run_local()
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_estimator/python/estimator/training.py", line 714, in run_local
    saving_listeners=saving_listeners)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_estimator/python/estimator/estimator.py", line 370, in train
    loss = self._train_model(input_fn, hooks, saving_listeners)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_estimator/python/estimator/estimator.py", line 1161, in _train_model
    return self._train_model_default(input_fn, hooks, saving_listeners)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_estimator/python/estimator/estimator.py", line 1195, in _train_model_default
    saving_listeners)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_estimator/python/estimator/estimator.py", line 1494, in _train_with_estimator_spec
    _, loss = mon_sess.run([estimator_spec.train_op, estimator_spec.loss])
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/monitored_session.py", line 754, in run
    run_metadata=run_metadata)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/monitored_session.py", line 1259, in run
    run_metadata=run_metadata)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/monitored_session.py", line 1360, in run
    raise six.reraise(*original_exc_info)
  File "/usr/local/lib/python3.6/dist-packages/six.py", line 693, in reraise
    raise value
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/monitored_session.py", line 1345, in run
    return self._sess.run(*args, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/monitored_session.py", line 1418, in run
    run_metadata=run_metadata)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/training/monitored_session.py", line 1176, in run
    return self._sess.run(*args, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/client/session.py", line 956, in run
    run_metadata_ptr)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/client/session.py", line 1180, in _run
    feed_dict_tensor, options, run_metadata)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/client/session.py", line 1359, in _do_run
    run_metadata)
  File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/client/session.py", line 1384, in _do_call
    raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InvalidArgumentError: 2 root error(s) found.
  (0) Invalid argument:  assertion failed: [[0.748][0.758]] [[0.67][0.67]]
     [[{{node Assert/AssertGuard/else/_123/Assert}}]]
     [[IteratorGetNext]]
  (1) Invalid argument:  assertion failed: [[0.748][0.758]] [[0.67][0.67]]
     [[{{node Assert/AssertGuard/else/_123/Assert}}]]
     [[IteratorGetNext]]
     [[IteratorGetNext/_8451]]
0 successful operations.
0 derived errors ignored.

Although I found similar questions having the similar title but the errors are not the same. Here additionally mentioning I am using tensorflow-gpu==1.15.0 and the model being used for fine tuning is ssd_mobilenet_v2_coco.

Any clues why this error is happening?

Answer 1

Alright! a formal answer might help. I fixed this issue in two steps. Firstly, while creating the CSV you should make sure there is no invalid entries. I mean, no invalid image and/or no image having bounding box(es) outside of the image, i.e, first check if the xmin, ymin, xmax, ymax all are within the image's resolution and they are not negatives. Also check the width and height are positives.

Secondly, while making tf_example I performed some additional checks to make sure all coordinates are still within the image. tfrecord wants the coordinates to be scaled in [0, 1]. Although, logically if we do the first step it is not necessary to check it again. But what I had found, probably due to some floating point precision problems these scaled coordinates sometimes got bigger than 1.0 or less than 0.0 and created this error again. So I some additional checks to the following function to make sure each of the entries are valid before I write them to tfrecord. In case if they are > 1.0 I made them 1.0 and if < 0.0 I made them 0.0. Following is the code:

def create_tf_example(group, path):
    with tf.gfile.GFile(os.path.join(path, '{}'.format(group.filename)), 'rb') as fid:
        encoded_jpg = fid.read()
    encoded_jpg_io = io.BytesIO(encoded_jpg)
    image = Image.open(encoded_jpg_io)
    width, height = image.size

    filename = group.filename.encode('utf8')
    image_format = b'jpg'
    xmins = []
    xmaxs = []
    ymins = []
    ymaxs = []
    classes_text = []
    classes = []

    for index, row in group.object.iterrows():
        
        ########### ADDITIONAL CHECKS START HERE ###################

        xmn = row['xmin'] / width
        if xmn < 0.0:
            xmn = 0.0
        elif xmn > 1.0:
            xmn = 1.0
        xmins.append(xmn)

        xmx = row['xmax'] / width
        if xmx < 0.0:
            xmx = 0.0
        elif xmx > 1.0:
            xmx = 1.0
        xmaxs.append(xmx)

        ymn = row['ymin'] / height
        if ymn < 0.0:
            ymn = 0.0
        elif ymn > 1.0:
            ymn = 1.0
        ymins.append(ymn)

        ymx = row['ymax'] / height
        if ymx < 0.0:
            ymx = 0.0
        elif ymx > 1.0:
            ymx = 1.0
        ymaxs.append(ymx)

        ############ ADDITIONAL CHECKS END HERE ####################

        classes_text.append(row['class'].encode('utf8'))
        classes.append(class_text_to_int(row['class']))

    tf_example = tf.train.Example(features=tf.train.Features(feature={
        'image/height': dataset_util.int64_feature(height),
        'image/width': dataset_util.int64_feature(width),
        'image/filename': dataset_util.bytes_feature(filename),
        'image/source_id': dataset_util.bytes_feature(filename),
        'image/encoded': dataset_util.bytes_feature(encoded_jpg),
        'image/format': dataset_util.bytes_feature(image_format),
        'image/object/bbox/xmin': dataset_util.float_list_feature(xmins),
        'image/object/bbox/xmax': dataset_util.float_list_feature(xmaxs),
        'image/object/bbox/ymin': dataset_util.float_list_feature(ymins),
        'image/object/bbox/ymax': dataset_util.float_list_feature(ymaxs),
        'image/object/class/text': dataset_util.bytes_list_feature(classes_text),
        'image/object/class/label': dataset_util.int64_list_feature(classes),
    }))
    return tf_example

Also there is another corner case which might be responsible for initiating this situation. It is related to how you annotated the bounding boxes. First I am describing the correct way of annotating. You will understand the rest yourself. While drawing those annotation boxes if you drag the mouse from left-top to right-bottom the annotator tool is considering the left-top point as the first point i.e, (xmin, ymin) and the right-bottom point as the second point i.e, (xmax, ymax). It is completely okay because in this case automatically the conditions xmin < xmax and ymin < ymax hold. But what happens when you do something different? Like, if you drag the mouse from right-bottom point to left-top point, therefore, the annotator tool might take the right-bottom point as (xmin, ymin) and the left-top point as (xmax, ymax). Which is completely wrong. As in this case xmax is becoming less than xmin and same problem occurs for ymax and ymin. So make sure your annotator software is capable of handling these kinds of situations by observing how you drag the mouse.

So, if you find the bounding boxes' annotations do have this problem, then you can easily correct the CSV by updating the values of xmin, xmax, ymin and ymax as followings:

import numpy as np

xmin_new = np.min(xmin, xmax)
xmax_new = np.max(xmin, xmax)
ymin_new = np.min(ymin, ymax)
ymax_new = np.max(ymin, ymax)

Also note that, I have used different variables to take the new values, replacing the values of the old variables (xmin, xmax, ymin, ymax) will make the further calculation wrong, as while taking np.max() or np.min() the expression expects the previous values of them, not the values we just updated during this process.

Answer 2

I had the exact same error when building my own tfrecords to retrain my model. The issue was that the height of one of the labeled boxes was negative. I'd recommend checking the sanity of your data.