| Unrecognized file_type: {} | validation | error | |
| When used with `multi_worker_mirrored`, valid values for all | validation | error | |
| When used with `mirrored`, valid values for all_reduce_alg a | validation | error | |
| `num_gpus` can not be negative. | validation | error | |
| distribution_strategy must be a string but got: %s. | validation | error | |
| When {num_gpus} GPUs are specified, distribution_strategy fl | validation | error | |
| `OneDeviceStrategy` can not be used for more than one device | validation | error | |
| Unrecognized Distribution Strategy: %r | validation | error | |
| Must specify task_index when number of workers > 1 | validation | error | |
| Optimizer has to be instance ofoptimization.ExponentialMovin | validation | error | |
| The loss value is NaN after training loop and it happens %d | exception | error | |
| The loss value is {loss_value}, which is larger than the bou | exception | error | |
| Coordinator uninitialized for async run. Call init_async() f | validation | error | |
| Stateful eval loop is not supported in async training. | validation | error | |
| The trainer requires the configuration contains an attribute | validation | error | |
| The trainer requires the configuration contains an attribute | validation | error | |
| The trainer requires the configuration contains an attribute | validation | error | |
| The trainer requires the configuration contains an attribute | validation | error | |
| If the function_keys is a list, it must contain a single ele | validation | error | |
| Failed to obtain a unique export directory name after {MAX_D | exception | error | |
| Unknown file_type: {file_type} | exception | error | |
| `input_path` should be either (1) a str indicating a file pa | exception | error | |
| %s does not match any files. | exception | error | |
| At most one of `input_path` and `tfds_name` can be specified | exception | error | |
| A combine_fn is required if `input_path` or `tfds_name` is a | exception | error | |
| `tfds_name` is %s, but `tfds_split` is not specified. | exception | error | |
| It is unexpected that `tfds_builder` is None and there is al | exception | error | |
| `matched_files` should be a list or dict. | exception | error | |
| Shared tf.data service does not support round-robin tf.data | exception | error | |
| Collection path {} at position {} already registered as a fu | exception | error | |
| Function or class {} registered multiple times. | exception | error | |
| collection path {entry_name} at position {h_idx} is never re | exception | error | |
| registration key {reg_key} is never registered. Please make | exception | error | |
| model_dir must be specified, but got None | exception | error | |
| The mode is not implemented: %s | exception | error | |
| Path not exist while traversing the dictionary: d with keys: | exception | error | |
| The value extracted with keys: %s is not a leaf of the dicti | exception | error | |
| best checkpoint metric comp must be one of higher, lower. Go | exception | error | |
| Differential privacy config is specified but task.create_opt | validation | error | |
| The flag --experiment must be specified. | validation | error | |
| model_dir must be specified, but got None | validation | error | |
| global_step not found in checkpoint {}. If you want to run f | validation | error | |
| We only support 'inner_group_num' and 'num_hidden_groups' as | validation | error | |
| Unsupported model_type %s. | validation | error | |
| %s does not match any files. | validation | error | |
| model_export_path must be specified. | validation | error | |
| model must be a tf_keras.Model object. | validation | error | |
| `input_fn` should be a closure that returns a dataset. | validation | error | |
| steps_per_loop should be positive integer. | validation | error | |
| only call `run_customized_training_loop()` with named argume | validation | error | |
| `strategy`, `model_fn`, `loss_fn`, `model_dir`, `steps_per_e | validation | error | |
| TPUStrategy should not run eagerly as it heavily relies on g | validation | error | |
| `eval_step` is required when `eval_input_fn ` is not none. | validation | error | |
| if `metric_fn` is specified, metric_fn must be a callable. | validation | error | |
| User should set optimizer attribute to model inside `model_f | validation | error | |
| sub_model_export_name is specified as %s, but sub_model is N | validation | error | |
| steps should be an Tensor. Python object may cause retracing | validation | error | |
| Export path is not specified: %s | validation | error | |
| Distribution strategy has not been specified. | validation | error | |
| Unsupported mode is specified: %s | validation | error | |
| Distribution strategy is not specified. | validation | error | |
| %s does not match any files. | validation | error | |
| `eval_squad` only supports one predict file, but got %s | validation | error | |
| Export path is not specified: %s | validation | error | |
| Invalid model instance: %s, it should be a %s | validation | error | |
| Model %s is not supported. | validation | error | |
| Parser %s is not supported. | validation | error | |
| mode is not defined. | validation | error | |
| Not implemented! | exception | error | |
| mode is not defined. | validation | error | |
| mode is not defined. | validation | error | |
| Please run accumulate() first | exception | error | |
| Missing the required key `{}` in predictions! | validation | error | |
| Missing the required key `{}` in groundtruths! | validation | error | |
| One and only one of `annotation_file` and `gt_dataset` needs | validation | error | |
| The `eval_type` can only be either `box` or `mask`. | validation | error | |
| Results do not correspond to the current dataset! | exception | error | |
| Evaluator %s is not supported. | exception | error | |
| steps_per_loop should be positive integer. | exception | error | |
| steps should be an Tensor. Python object may cause retracing | exception | error | |
| if `train_metric_fn` is specified, train_metric_fn must be a | exception | error | |
| if `eval_metric_fn` is specified, eval_metric_fn must be a c | exception | error | |
| User should set optimizer attribute to model inside `model_f | exception | error | |
| total loss is NaN. | exception | error | |
| model_dir must be set. | exception | error | |
| checkpoint path is empty | exception | error | |
| `strategy` should not be None. You need to specify `strategy | exception | error | |
| checkpoint_path cannot be empty. | exception | error | |
| Mode not found: %s. | exception | error | |
| Must provide at least one of training_file_pattern and eval_ | exception | error | |
| Backbone model `{}` is not supported. | exception | error | |
| The multi-level feature model `{}` is not supported. | exception | error | |
| Unsupported activation `{}`. | exception | error | |
| The minimum backbone level %d should be less or equal to FPN | exception | error | |
| Unsupported activation `{}`. | exception | error | |
| Unsupported activation `{}`. | exception | error | |
| Unsupported activation `{}`. | exception | error | |
| Unsupported activation `{}`. | exception | error | |
| The resnet_depth should be in [%s]. Not a valid resnet_depth | exception | error | |
| Activation {} not implemented. | exception | error | |
| Duplicate feats found for output level {}. | exception | error | |
| Output level is out of range [{}, {}] | exception | error | |
| SpineNet {} is not a valid architecture. | exception | error | |
| Unimplemented eval_metrics | exception | error | |
| Variables to load is empty. | exception | error | |
| Model %s is not supported. | exception | error | |
| Unsupported learning rate type: {}. | exception | error | |
| build_loss_fn() must be called after build_model(). | exception | error | |
| "%s" is missing in outputs, requried %s found %s | exception | error | |
| build_loss_fn() must be called after build_model(). | exception | error | |
| Unsupported optimizer type `{}`. | exception | error | |
| build_loss_fn() must be called after build_model(). | exception | error | |
| "%s" is missing in outputs, requried %s found %s | exception | error | |
| build_loss_fn() must be called after build_model(). | exception | error | |
| "{}" is missing in outputs, requried {} found {} | exception | error | |
| "{}" is missing in labels, requried {} found {} | exception | error | |
| boxes.shape[-1] is {:d}, but must be 4. | exception | error | |
| scale is {}, but outer box scale must be greater than 1.0. | exception | error | |
| encoded_boxes.shape[-1] is {:d}, but must be 4. | exception | error | |
| encoded_boxes_lrtb.shape[-1] is {:d}, but must be 4. | exception | error | |
| boxes.shape[1] is {:d}, but must be 4. | exception | error | |
| Invalid split name {}!!! | exception | error | |
| translations rank must be statically known | exception | error | |
| translations should have rank 1 or 2. | exception | error | |
| Angles should have a rank 0 or 1. | exception | error | |
| output_shape must be a 1-D Tensor of 2 elements: new_height, | exception | error | |
| Invalid augmentation_name: {} | exception | error | |
| AverageModelCheckpoint is only used when trainingwith Moving | exception | error | |
| {} is not a valid mode. | exception | error | |
| Invalid dataset received. Received: {}. Supported datasets i | exception | error | |
| Invalid model received. Received: {}. Supported models for{} | exception | error | |
| Invalid DType provided. Supported types: {} | exception | error | |
| Unknown builder type {} | exception | error | |
| Dataset must specify a path for the data files. | exception | error | |
| The builder does not support a global batch size with more t | exception | error | |
| The tf_data_service flag requires Tensorflow version >= 2.3. | exception | error | |
| TpuBatchNormalization does not support fused=True. | exception | error | |
| num_shards: %d mod shards_per_group: %d, should be 0 | exception | error | |
| Unknown model name {} | exception | error | |
| optimizer is not an object of tf_keras.optimizers.Optimizer | exception | error | |
| Unknown optimizer %s | exception | error | |
| `model` must be provided if using `ExponentialMovingAverage` | exception | error | |
| Input must be of size [height, width, C>0] | exception | error | |
| len(means) must match the number of channels | exception | error | |
| len(stddev) must match the number of channels | exception | error | |
| The length of boundaries must be 1 less than the length of m | exception | error | |
| Input must be of size [height, width, C>0] | exception | error | |
| len(means) must match the number of channels | exception | error | |
| Batch size must be divisible by number of replicas : {} | exception | error | |
| Hidden size ({}) must be divisible by the number of heads ({ | exception | error | |
| Reference and translation files have different number of lin | exception | error | |
| Download/extraction failed for url %s to path %s | exception | error | |
| mode {} is not valid. | exception | error | |
| Not valid params: param_set={} num_gpus={} | exception | error | |
| Padded decoding on CPU/GPUs is not supported. | exception | error | |
| Custom training loop on GPUs is not implemented. | exception | error | |
| Keras model.fit on TPUs is not implemented. | exception | error | |
| Invalid mode {} | exception | error | |
| File output is a directory, will not save outputs to file. | exception | error | |
| Batch size must be divisible by number of replicas : {} | exception | error | |
| Do not have sample info. | exception | error | |
| Task not found: %s | exception | error | |
| Unsupported string type: %s | exception | error | |
| For training, each question should have exactly 1 answer. | exception | error | |
| `train_input_fn`, `total_training_steps`, `steps_per_loop`, | exception | error | |
| Model directory must be specified. | exception | error | |
| steps should be an Tensor. Python object may cause retracing | exception | error | |
| Initializer {} not supported | exception | error | |
| Unsupported attention type: {} | exception | error | |
| Unknown `attn_type` {}. | exception | error | |
| Invalid summary type provided: %s | exception | error | |
| Only support expanding the first or the last dimension. Got: | exception | error | |
| Shape not supported, {}, {} | exception | error | |
| The mode is not implemented: %s | exception | error | |
| Please do not set train_dataset. Progressive training relies | exception | error | |
| Please do not set eval_dataset. Progressive training relies | exception | error | |
| No gradients provided for any variable: %s. | exception | error | |
| The bind decorator is supposed to apply on the class attribu | exception | error | |
| Inside a program, we should not bind the config with a class | exception | error | |
| The `BUILDER` type is not supported: {builder} | exception | error | |
| Subconfig_type should be subclass of ParamsDict, found {!r} | exception | error | |
| Invalid sequence: only supports single level {!r} of {!r} or | exception | error | |
| Unknown type: {!r} | exception | error | |
| dict value not supported in converting. | exception | error | |
| `BUILDER` is a property and `_BUILDER` is the reserved class | exception | error | |
| The Config has been locked. No change is allowed. | exception | error | |
| The key {!r} is internally reserved. Can not be overridden. | exception | error | |
| The key {!r} does not exist in {!r}. To extend the existing | exception | error | |
| type: {!r} is not a valid key! | exception | error | |
| The key `%{}` does not exist. To extend the existing keys, u | exception | error | |
| The ParamsDict has been locked. No change is allowed. | exception | error | |
| The key `{}` does not exist. | exception | error | |
| The key `{}` is reserved. No change is allowes. | exception | error | |
| The key `%{}` is internally reserved. Can not be overridden. | exception | error | |
| The key `{}` does not exist. To extend the existing keys, us | exception | error | |
| Only support binary relation in restriction. | exception | error | |
| Found inconsistency between key `{}` and key `{}`. | exception | error | |
| Unsupported relation in restriction. | exception | error | |
| Malformed hyperparameter value while parsing CSV string: %s | exception | error | |
| Did not pass all values of array: %s | exception | error | |
| Unknown input type to parse. | exception | error | |
| _instantiate_sub_task_models() is not implemented. | exception | error | |
| Duplicated tasks found, task.name is %s | exception | error | |
| The tasks argument has an invalid type: %s | exception | error | |
| The iterator output is neither a tuple nor a dictionary. It | exception | error | |
| Task sampler type not supported | exception | error | |
| The mode is not implemented: %s | exception | error | |
| Swapping weights must occur under a tf.distribute.Strategy. | exception | error | |
| Applying sparse gradients is not implemented. | exception | error | |
| Expect non empty {self.values} | exception | error | |
| Boundaries length is equal to learning rate levels length{le | exception | error | |
| %s already registered in LEGACY_OPTIMIZERS_CLS. | exception | error | |
| %s already registered in NEW_OPTIMIZERS_CLS. | exception | error | |
| Optimizer type must be specified | exception | error | |
| Learning rate type must be specified | exception | error | |
| `decay` is deprecated in new Keras optimizer, please reflect | exception | error | |
| EMA can only work with the legacy optimizer, please set `use | exception | error | |
| OptimizerFactory.build_optimizer returning a non-optimizer o | exception | error | |
| Unexpected dtype: %s | exception | error | |
| For the tensor `%s`, the actual tensor rank `%d` (shape = %s | exception | error | |
| Unknown replica context. The `get_replica_id` method relies | exception | error | |
| {value} has unknown batch. | exception | error | |
| The BUILDER returns an unexpected instance. The `build_encod | exception | error | |
| ExperimentConfig.task.init_checkpoint must be a directory fo | exception | error | |
| Invalid `mnli_type`: %s | exception | error | |
| language %s is not supported for PAWS-X task. | exception | error | |
| Split {} not available. | exception | error | |
| language %s is not supported for XNLI task. | exception | error | |
| Task not found: %s | exception | error | |
| No data processor found for the given regression task. | exception | error | |
| Unsupported tokenization: %s | exception | error | |
| FLAG vocab_file for word-piece tokenizer is not specified. | exception | error | |
| FLAG sp_model_file for sentence-piece tokenizer is not speci | exception | error | |
| cannot use ngram masking without whole word masking | exception | error | |
| Must specify either `tfds_name` and `tfds_split` or `input_p | exception | error | |
| Must specify exactly one of vocab_file (with matching lower_ | exception | error | |
| `reuse_length` and `seq_length` should both be a multiple of | exception | error | |
| `seq_length` should be a multiple of `permutation_size`. | exception | error | |
| `max_predictions_per_seq` must be set. | exception | error | |
| `boundary` must be provided for {} strategy | exception | error | |
| Invalid sample strategy. | exception | error | |
| The seq_bucket_lengths cannot be empty. | exception | error | |
| Currently there is no support for more than text field while | exception | error | |
| Unsupported tokenization: %s | exception | error | |
| `tfds_name` and `tfds_split` should be specified or unspecif | exception | error | |
| Must specify either `tfds_name` and `tfds_split` or `input_p | exception | error | |
| Unexpected empty text fields. | exception | error | |
| Must specify exactly one of vocab_file (with matching lower_ | exception | error | |
| For training, each question should have exactly 1 answer. | exception | error | |
| For training, each question should have exactly 1 answer. | exception | error | |
| Unexpected negative label_id: %s | exception | error | |
| The token budget, global batch size, is too small to yield 0 | exception | error | |
| Unexpected that both %s and %s are in config. | exception | error | |
| Unsupported `label_type`. Given: %s, expected `int` or `floa | exception | error | |
| Flag `%s` at %s does not exist. | exception | error | |
| Flag `%s` must be provided in mode %s. | exception | error | |
| When `hub_module_url` is specified, `init_checkpoint` and `m | exception | error | |
| Both `init_checkpoint` and `model_config_file` should be spe | exception | error | |
| %s does not exist. | exception | error | |
| Task %s not supported. | exception | error | |
| Flag `%s` at %s does not exist. | exception | error | |
| Flag `%s` must be provided in mode %s. | exception | error | |
| When `hub_module_url` is specified, `init_checkpoint` and `m | exception | error | |
| Both `init_checkpoint` and `model_config_file` should be spe | exception | error | |
| %s does not exist. | exception | error | |
| Task %s not supported. | exception | error | |
| Reference and translation files have different number of lin | exception | error | |
| Intermediate_size (%d) isn't a multiple of num_blocks (%d). | exception | error | |
| src_block_size must be specified. | exception | error | |
| num_kv_heads must be 1. Grouped-query attention is not suppo | exception | error | |
| sigmoid_attn_bias must be specified for sigmoid attn. | exception | error | |
| query_shape[-2] must be divisible by src_block_size. | exception | error | |
| key_shape[-2] must be divisible by tgt_block_size. | exception | error | |
| src_num_blocks must be equal to tgt_num_blocks. | exception | error | |
| use_causal_mask is not supported. | exception | error | |
| The dropout_position should be either `before_residual` or ` | exception | error | |
| "likelihood" must be one of {_SUPPORTED_LIKELIHOOD}, got {li | exception | error | |
| "logits" cannot be None when likelihood={self.likelihood} | exception | error | |
| likelihood={self.likelihood} only support univariate logits. | exception | error | |
| Illegal padding value; must be one of "left", "right" or Non | exception | error | |
| window_decay should be in (0.0, 1.0) and not None. | exception | error | |
| Unsupported feature_transform. The supported feature_transfo | exception | error | |
| There is nothing to redraw when num_random_features <= 0. | exception | error | |
| use_causal_windowed and short_seq methods are mutually exclu | exception | error | |
| Cache is not supported when training is True. | exception | error | |
| Cache is not supported for non use_causal_windowed case. | exception | error | |
| Cache is not supported when begin_kernel is set since the ba | exception | error | |
| Cache is not supported for feature_transform %s | exception | error | |
| Unknown `output` value "%s". `output` can be either "logits" | exception | error | |
| MaskedLM cannot be directly serialized because it has variab | exception | error | |
| Abstract method | exception | error | |
| Only "no_norm" and "layer_norm" and supported. | exception | error | |
| The bottleneck size {intra_bottleneck_size} is not a multipl | exception | error | |
| The width of the input tensor {input_width} != hidden size { | exception | error | |
| Unknown `output` value "%s". `output` can be either "logits" | exception | error | |
| hidden size %d cannot be smaller than embedding width %d. | exception | error | |
| MaskedLM cannot be directly serialized because it has variab | exception | error | |
| Router is an abstract class that should be subclassed. | exception | error | |
| MaskedRouter is an abstract class that should be subclassed. | exception | error | |
| Input shape {input_shape} is inconsistent with num_experts { | exception | error | |
| examples_per_group={self._examples_per_group} is larger than | exception | error | |
| StridedTransformerEncoderBlock does not support `output_rang | exception | error | |
| StridedTransformerEncoderBlock does not support block sparse | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| {self.__class__} does not support `output_range` argument. | exception | error | |
| {self.__class__} does not support block sparse attention. | exception | error | |
| Unexpected inputs to {self.__class__} with length at {len(in | exception | error | |
| `max_length` must be an Integer, not `None`. | exception | error | |
| If inputs is None, `length` must be set in RelativePositionE | exception | error | |
| If inputs is not None, `length` must equal to input_shape[1] | exception | error | |
| reuse_attention should be between -1 and %d in call to %s. | exception | error | |
| reuse_attention_scores cannot be None when reuse_attention i | exception | error | |
| The type of input shape argument is not supported, got: %s | exception | error | |
| The input size (%d) is not a multiple of the number of atten | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| reuse_attention_scores cannot be None when reuse_attention ! | exception | error | |
| Linformer and Sigmoid attention are not supported in ReZero | exception | error | |
| Block sparse attention does not support Multi-query attentio | exception | error | |
| The type of input shape argument is not supported, got: %s | exception | error | |
| TransformerLayer expects a three-dimensional input of shape | exception | error | |
| When passing a mask tensor to TransformerLayer, the mask ten | exception | error | |
| The input size (%d) is not a multiple of the number of atten | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| `layer` must be a `tf_keras.layer.Layer`. Observed `{}` | exception | error | |
| layer must be a `tf_keras.layer.Conv2D` instance. You passed | exception | error | |
| import tensorflow_text failed, please install 'tensorflow-te | exception | error | |
| TODO(b/170480226): implement | exception | error | |
| Exact one of `model_file_path` and `model_serialized_proto` | exception | error | |
| `tokenize_with_offsets` is not supported when `strip_diacrit | exception | error | |
| `tokenize_with_offsets` is not supported yet when `strip_dia | exception | error | |
| Only 'round_robin' and 'waterfall' algorithms are supported, | exception | error | |
| At least one input is required for packing | exception | error | |
| All inputs for packing must have the same known rank, found | exception | error | |
| Unsupported truncator: %s | exception | error | |
| Not implemented yet. | exception | error | |
| The last dimension of the inputs to `TNExpandCondense` shoul | exception | error | |
| TNTransformerExpandCondense expects a three-dimensional inpu | exception | error | |
| When passing a mask tensor to TNTransformerExpandCondense, t | exception | error | |
| The input size (%d) is not a multiple of the number of atten | exception | error | |
| TransformerLayer expects a three-dimensional input of shape | exception | error | |
| The hidden size (%d) is not a multiple of the number of atte | exception | error | |
| TransformerDecoderBlock must have 5 inputs, when it uses mul | exception | error | |
| TransformerDecoderBlock must have 4 inputs, but it got: %d | exception | error | |
| Block sparse attention only supports Multi-query attention.P | exception | error | |
| Setting `diff_q_and_kv_attention_layer_norm` to Truewhen `no | exception | error | |
| The type of input shape argument is not supported, got: %s | exception | error | |
| Block sparse attention does not support talking heads. Pleas | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| Unexpected keys in input dictionary to: {inputs.keys()} | exception | error | |
| Missing required key `input_tensor` in input dictionary. | exception | error | |
| The type of input shape argument is not supported, got: %s | exception | error | |
| TransformerScaffold expects a three-dimensional input of sha | exception | error | |
| The input size (%d) is not a multiple of the number of atten | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| TransformerLayer expects a three-dimensional input of shape | exception | error | |
| When passing a mask tensor to TransformerXLBlock, the mask t | exception | error | |
| The input size (%d) is not a multiple of the number of atten | exception | error | |
| Weight and label tensors were not of the same rank. weights. | exception | error | |
| Weighted sparse categorical crossentropy expects `labels` to | exception | error | |
| The passed network's output length is %s, which is less than | exception | error | |
| Classification heads should have unique names. | exception | error | |
| encoder_network's output should be either a list or a dict, | exception | error | |
| Unknown `output` value "%s". `output` can be either "logits" | exception | error | |
| output type %s is not supported | exception | error | |
| The call method expects either `inputs` or `embedded_inputs` | exception | error | |
| When not using a bias the b_init must be None. | exception | error | |
| Invalid activation function string is passed: %s | exception | error | |
| At least one of inputs and dense_inputs must not be None. | exception | error | |
| At least one of encoder_input_tokens and encoder_dense_input | exception | error | |
| Invalid summary type provided: %s. | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| Unknown `output` value "%s". `output` can be either "logits" | exception | error | |
| When input hidden_cls to EncoderScaffold %s is a list, it mu | exception | error | |
| The EncoderScaffold %s does not have a reference to the embe | exception | error | |
| The EncoderScaffold %s does not have a reference to the embe | exception | error | |
| Unexpected inputs type (%s) to %s. | exception | error | |
| Unsupported mixing mechanism: %s | exception | error | |
| The lengths of strides and axes need to match. | exception | error | |
| Lengths of pool_stride and num_layers are not equal. | exception | error | |
| Fractional pooling is only supported for `pool_type`=`trunca | exception | error | |
| unpool_length is not supported by truncated_avg now. | exception | error | |
| pool_type not supported. | exception | error | |
| Unexpected inputs to %s with length at %d. | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| unpool_length is not supported by append_dense_inputs now. | exception | error | |
| If `use_dynamic_slicing` is True, `max_sequence_length` must | exception | error | |
| PositionEmbedding expects a 3-dimensional input tensor of sh | exception | error | |
| PositionEmbedding must have `use_dynamic_slicing` set to Tru | exception | error | |
| When `use_dynamic_slicing` is False, max_sequence_length sho | exception | error | |
| Unknown `output` value "%s". `output` can be either "logits" | exception | error | |
| `start_n_top` must be greater than 1. | exception | error | |
| Unexpected inputs type (%s) to %s. | exception | error | |
| Unsupported mixing mechanism: %s | exception | error | |
| Unknown `attention_type` {}. | exception | error | |
| `masked_tokens` must be provided in order to initialize the | exception | error | |
| Invalid dtype: %s | exception | error | |
| Expect rank for tensor shape, but got None. | exception | error | |
| Expect rank for target shape, but got None. | exception | error | |
| initial_cache element for key '%s' has dtype %s that does no | exception | error | |
| Expect rank for tensor shape, but got None. | exception | error | |
| Expect rank for target shape, but got None. | exception | error | |
| Invalid dtype: %s | exception | error | |
| initial_cache element for key '%s' has dtype %s that does no | exception | error | |
| Unsupported optimizer type: | exception | error | |
| Failed to identify the task class. The provided task name is | exception | error | |
| Invalid function key for the module: %s with key %s. Valid k | exception | error | |
| EOS token not in tokenizer vocab.Please make sure the tokeni | exception | error | |
| At most one of `hub_module_url` and `init_checkpoint` can be | exception | error | |
| Unsupported tokenization method: {} | exception | error | |
| At most one of `hub_module_url` and `init_checkpoint` can be | exception | error | |
| You must specify a temporary directory, either in params.inp | exception | error | |
| Unexpected tokenization: %s | exception | error | |
| Invalid metric_type: {} | exception | error | |
| At most one of `hub_module_url` and `init_checkpoint` can be | exception | error | |
| At most one of `hub_module_url` and `init_checkpoint` can be | exception | error | |
| EOS token not in tokenizer vocab.Please make sure the tokeni | exception | error | |
| Setencepiece model path not provided. | exception | error | |
| Exactly one of `vocab_file` and `sp_model_file` can be speci | exception | error | |
| Exactly one of `bert_config_file` and `encoder_config_file` | exception | error | |
| Must set vocab_file or sp_model_file. | exception | error | |
| Exactly one of `bert_config` and `encoder_config` can be spe | exception | error | |
| Must pass do_lower_case if passing vocab_file. | exception | error | |
| Must set vocab_file or sp_model_file | exception | error | |
| Must set exactly one of vocab_file, sp_model_file | exception | error | |
| Failed to clean up TemporaryDirectory | exception | error | |
| Internal tool error: failed to suppress {} Assert ops in Sav | exception | error | |
| Unsupported converted_model: %s | exception | error | |
| Unsupported converted_model: %s | exception | error | |
| You passed in `--do_lower_case=%s` with `--init_checkpoint=% | exception | error | |
| Unsupported string type: %s | exception | error | |
| strides > 1 requires use_projections=True, otherwise the inp | exception | error | |
| Expect input spec to be 4 or 5 dimensions {input_specs.shape | exception | error | |
| Not a valid assemblenet_depth: | exception | error | |
| Expect input spec to be 4 or 5 dimensions {input_specs.shape | exception | error | |
| Not a valid assemblenet_depth: | exception | error | |
| Rank of Tensor %s must be known | exception | error | |
| _force_data_dependency only supports floating dtypes. | exception | error | |
| Found unexpected kwargs for `grad`: | exception | error | |
| `true_fn` must be callable. | exception | error | |
| `false_fn` must be callable. | exception | error | |
| RecomputeContext is required when force_recomputation=True. | exception | error | |
| x has to be a floating point tensor since it's going to be | exception | error | |
| rate must be a scalar tensor or a float in the range [0, 1), | exception | error | |
| Detection with center point is only available | exception | error | |
| Unsupported datatype used in parser only {float16, bfloat16, | exception | error | |
| filter size and residual block repetition lists must have th | exception | error | |
| Invalid dimensions for box data: {} | exception | error | |
| Invalid tensor type: should be tf.float32 | exception | error | |
| Invalid dimensions for box data. | exception | error | |
| field %s does not exist | exception | error | |
| boxlist must contain all specified fields | exception | error | |
| indices should have rank 1 | exception | error | |
| indices should be an int32 / int64 tensor | exception | error | |
| boxlist must contain all specified fields | exception | error | |
| Image should be 3D tensor | exception | error | |
| Number of channels must be equal to the length of per-channe | exception | error | |
| Unsupported datatype used in ground truth builder only {floa | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| Only 'all' or 'backbone' can be used to initialize the model | exception | error | |
| Can't evaluate using annotation file when TFDS is used. | exception | error | |
| Number of blocks in temporal specs should equal to resnet_sp | exception | error | |
| Block fn `{}` is not supported. | exception | error | |
| Stem type {stem_type} not supported. | exception | error | |
| Number of elements in temporal_kernel_sizes must equal to bl | exception | error | |
| Should use the same batch normalization type. | exception | error | |
| context_level should be specified as odd number. | exception | error | |
| Expected features is a rank-5 tensor. Got shape %s | exception | error | |
| Padding for kernel size {} not known. | exception | error | |
| Unknown layer - {} | exception | error | |
| Unknown inner block type. | exception | error | |
| Unknown ConvNet variant - {} | exception | error | |
| Detection module not implemented for {} model. | exception | error | |
| Can only use image and boxes input for DeepMaskRCNNModel, Fo | exception | error | |
| Number of embedding features (num_pos_features) must be even | exception | error | |
| hidden_size must be a multiple of 2. | exception | error | |
| The type of input shape argument is not supported, got: %s | exception | error | |
| The input size (%d) is not a multiple of the number of atten | exception | error | |
| TransformerLayer expects a three-dimensional input of shape | exception | error | |
| The hidden size (%d) is not a multiple of the number of atte | exception | error | |
| For the tensor `%s`, the actual tensor rank `%d` (shape = %s | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| Number of transformer layer must be equal or divisible. | exception | error | |
| Unknown distillation mode: {self.mode}. | exception | error | |
| Training mode has to be LAYER-WISE or END2END. | exception | error | |
| Can not find embedding layer in the encoder. | exception | error | |
| encoder_network's output should be either a list or a dict, | exception | error | |
| `teacher_model_init_checkpoint` is not specified. | exception | error | |
| Unsupported mode, only support `train` | exception | error | |
| Unsupported model/id type {backbone_cfg.model_id}. | exception | error | |
| TpuBatchNormalization does not support fused=True. | exception | error | |
| num_shards: %d mod shards_per_group: %d, should be 0 | exception | error | |
| Unsupported checkpoint format {checkpoint_format}. | exception | error | |
| Number of groups {groups} should be greater than 1 and less | exception | error | |
| GroupConv2D expects input to be in channels_last format. | exception | error | |
| Valid padding options are : same, or valid. | exception | error | |
| batch_norm_layer is not a class. | exception | error | |
| Number of input channels: {input_channel} are not divisible | exception | error | |
| Number of groups should be greater than 1 and less than the | exception | error | |
| Number of input channels: {input_shape[-1]} are not divisibl | exception | error | |
| Please specify epsilon for unknown input data type | exception | error | |
| Unsupported act_type {} | exception | error | |
| Unknown conv type: {} | exception | error | |
| Upsampling type {} is not supported. | exception | error | |
| Unsupported pooling type {}. | exception | error | |
| Incompatible Resampling : feat shape {}x{} target_shape: {}x | exception | error | |
| unknown weight_method %s | exception | error | |
| Only bifpn config is supported. | exception | error | |
| Only bifpn config is supported | exception | error | |
| Unknown model name {} | exception | error | |
| Unknown model name {model_config_name}. Only supportmodel co | exception | error | |
| Received skip type within block creation. | exception | error | |
| Unsupported IBN type {block_op_type.type}. | exception | error | |
| Number of filters: {conv_filters} is not divisible by size o | exception | error | |
| Fused projection is not supported. | exception | error | |
| Cannot finalize with {finalize_method[i]}. | exception | error | |
| Must provide a representative dataset when quantizing the mo | exception | error | |
| Saved model path is invalid. | exception | error | |
| If using MobileNet-EdgeTPU-Search model, pleasespecify the s | exception | error | |
| Model has to be mobilenet-edgetpu model or searchedmodel wit | exception | error | |
| Only 'all' or 'backbone' can be used to initialize the model | exception | error | |
| TFDS {} is not supported | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| Invalid metric_type: {} | exception | error | |
| At most one of `hub_module_url` and `init_checkpoint` can be | exception | error | |
| Unexpected inputs type to {self.__class__}. | exception | error | |
| The type of input shape argument is not supported, got: {typ | exception | error | |
| The input size ({hidden_size}) is not a multiple of the numb | exception | error | |
| Unexpected inputs to {self.__class__} with length at {len(in | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| Setting `diff_q_and_kv_attention_layer_norm` to Truewhen `no | exception | error | |
| The type of input shape argument is not supported, got: %s | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| Invalid metric_type: {} | exception | error | |
| At most one of `hub_module_url` and `init_checkpoint` can be | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| Number of embedding features (num_pos_features) must be even | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| Augmentation policy {} not supported. | exception | error | |
| segmentation_groundtruth_padded_size ([height, width]) needs | exception | error | |
| Augmentation policy {} not supported. | exception | error | |
| segmentation_groundtruth_padded_size ([height, width]) needs | exception | error | |
| Augmentation policy {} not supported. | exception | error | |
| Fusion type {} not supported. | exception | error | |
| Backbone min level should be less or equal to FPN min level | exception | error | |
| Inconsistent decoder type {decoder_type}. Need to be `maskco | exception | error | |
| Only support dictionary decoder_output. | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| Unsupported act_fn {} | exception | error | |
| Unsupported act_fn %s. | exception | error | |
| Unsurpported pool_type %s | exception | error | |
| Unsupport shape {} | exception | error | |
| Invalid sequence length: {length} or shape: ({height, width} | exception | error | |
| Sequence length: %s violates input size: (%s, %s). | exception | error | |
| Does not support relative attention for query shape: %s. | exception | error | |
| rel_attn_type {self.rel_attn_type} not implemented yet. | exception | error | |
| Unsupported input shape: {input_shape.as_list()}. | exception | error | |
| Unsupported norm_type {norm_type}. | exception | error | |
| Unsurpported pool_type {pool_type} | exception | error | |
| Feature map sizes {(h, w)} not divisible by window size ({wi | exception | error | |
| Feature map sizes {(h, w)} not divisible by window size ({gr | exception | error | |
| Unsupported norm_type {self._norm_type}. | exception | error | |
| Unsupported block_type {self._block_type[i]} | exception | error | |
| The number of `transfer_teacher_layers` %s does not match th | exception | error | |
| `transfer_teacher_layers` is not specified, and the number o | exception | error | |
| distill_ground_truth_ratio has to be within [0, 1]. | exception | error | |
| `teacher_model_init_checkpoint` is not specified. | exception | error | |
| The number of output groups must match #kernels. | exception | error | |
| The number of input channels must be divisible by the number | exception | error | |
| The number of pooling bins must be smaller than input sizes. | exception | error | |
| The number of Decoder inputs and settings must match. | exception | error | |
| A stage merge style in MOSAIC Decoder can only be concat_mer | exception | error | |
| Additional classifier layer is needed if final decoder proje | exception | error | |
| Only support MosaicDecoderHead for head. | exception | error | |
| Only support MosaicEncoderBlock for encoder. | exception | error | |
| The number of Decoder inputs and settings must match. | exception | error | |
| A stage merge style in MOSAIC Decoder can only be concat_mer | exception | error | |
| Additional classifier layer is needed if final decoder proje | exception | error | |
| The number of input channels must be divisible by the number | exception | error | |
| The number of pooling bins must be smaller than input sizes. | exception | error | |
| Export module for {type(params.task)} is not supported. | exception | error | |
| Unknown conv type: {} | exception | error | |
| Unknown squeeze excitation type: {} | exception | error | |
| External states should be used with causal mode. | exception | error | |
| Expected first spec to be StemSpec, got {} | exception | error | |
| Expected final spec to be HeadSpec, got {} | exception | error | |
| Lengths of block parameters differ: {}, {}, {} | exception | error | |
| Unknown block type {} | exception | error | |
| Strides must match in the spatial dimensions, got {} | exception | error | |
| Inconsistent backbone type {backbone_type} | exception | error | |
| The `{name}` argument must be a tuple of {size} integers. Re | exception | error | |
| The `{name}` argument must be a tuple of {size} integers. Re | exception | error | |
| Temporal conv with spatial kernel is not supported. | exception | error | |
| Unknown Squeeze Excitation type {} | exception | error | |
| %s average_pooling_type is not supported. | exception | error | |
| Expected input and output states to be the same. Got extra s | validation | error | |
| Got mismatched input and output state shapes: {} | validation | error | |
| Bad conversion, model outputs do not match. | validation | error | |
| Unrecongized dtype: {dtype} | validation | error | |
| Invalid attention bias type: %s | validation | error | |
| If `use_type_embeddings` is True, then `token_type_vocab_siz | validation | error | |
| Batch size must be divisible by number of replicas : {} | validation | error | |
| bert_layer should be built. | validation | error | |
| decoder_layer should be built. | validation | error | |
| Invalid call mode: %s | validation | error | |
| The model type is not defined: %s | exception | error | |
| data_folder must be set as the downloaded folder path. | validation | error | |
| vocab must be set as the filepath of BERT vocabulary. | validation | error | |
| Source encoder layers with %d objects does not match destina | validation | error | |
| groundtruth_padded_size ([height, width]) needs to bespecifi | validation | error | |
| Augmentation policy {} not supported. | validation | error | |
| segmentation_groundtruth_padded_size ([height, width]) needs | validation | error | |
| Length of class_weights should be {} | validation | error | |
| Expected the input masks (..., height, width) has rank >= 2, | validation | error | |
| `mask_head` needs to be provided for Panoptic Mask R-CNN. | validation | error | |
| `segmentation_decoder` needs to be provided for Panoptic Mas | validation | error | |
| Unsupported model type: %s | validation | error | |
| batch_size cannot be None for panoptic segmentation model. | validation | error | |
| PanopticSegmentationModule module not implemented for {} mod | validation | error | |
| batch_size cannot be None for panoptic segmentation model. | validation | error | |
| PanopticSegmentationModule module not implemented for {} mod | validation | error | |
| Unknown decoder type: {}! | validation | error | |
| Only 'all', 'backbone', 'decoder', 'segmentation_backbone' a | validation | error | |
| Unknown decoder type: {}! | validation | error | |
| `include_panoptic_masks` should be set to True when computin | validation | error | |
| `inputs` must be a sequence. | validation | error | |
| `inputs` must have two elements. | validation | error | |
| Unknown value {shape_for_attn} for shape_for_attention. | validation | error | |
| if `sequence_output` is not in encoder output, `latent_outpu | validation | error | |
| Unexpected inputs type to {self.__class__}. | validation | error | |
| decoder's output should be a dict,but got {decoder_output} | validation | error | |
| If `sequence_output` is not in encoder output,`latent_output | validation | error | |
| `sequence_output` must be in decoder output. | validation | error | |
| if `sequence_output` is not in encoder output, `latent_outpu | validation | error | |
| Unexpected inputs type to {self.__class__}. | exception | error | |
| decoder's output should be a dict,but got {decoder_output} | exception | error | |
| If `sequence_output` is not in encoder output,`latent_output | exception | error | |
| Unexpected inputs type to {self.__class__}. | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| Unknown pos encoding %s | exception | error | |
| Number of embedding features (num_pos_features) must be even | exception | error | |
| hidden_size must be a multiple of 2. | exception | error | |
| The length of encoded_feature_dropout_rates must be equal to | exception | error | |
| drop_rate {} is outside [0, 1) | exception | error | |
| init_checkpoint_modules=backbone is no longer supported. Spe | exception | error | |
| Unsupported init_checkpoint_modules: {self._task_config.init | exception | error | |
| A global init_checkpoint and a backbone init_checkpoint cann | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| Unknown order {} | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| The min_level must be >= 1, but {} found. | exception | error | |
| The min_level should be >= decoder output level, but {} < {} | exception | error | |
| Unsupported head type: {} | exception | error | |
| PointPillars model needs attribute heads. | exception | error | |
| Image_shape should not be None for evaluation. | exception | error | |
| An NMS algorithm is required for detection generator | exception | error | |
| Unrecognized file_type: {} | exception | error | |
| global_batch_size {} is not a multiple of num_replicas {} | exception | error | |
| Model num_classes must be 2 when not for all classes. | exception | error | |
| Attribute type {head.type} not supported. | exception | error | |
| Unrecognized input_type: {} | exception | error | |
| Shape of array should be {}, but {} found | exception | error | |
| Only "channels_last" mode is supported | exception | error | |
| {} does not have enough elements to gather, {} < {} | exception | error | |
| The length of attribute_heads should be 1 or 3, found {} | exception | error | |
| Only PolynomialDecay and ConstantSparsity are currently supp | exception | error | |
| Only "no_norm" and "layer_norm" are supported. | exception | error | |
| The bottleneck size {intra_bottleneck_size} is not a multipl | exception | error | |
| The width of the input tensor {input_width} != hidden size { | exception | error | |
| The type of input shape argument is not supported, got: %s | exception | error | |
| TransformerEncoderBlock expects a three-dimensional input of | exception | error | |
| The input size (%d) is not a multiple of the number of atten | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| Unknown `output` value "%s". `output` can be either "logits" | exception | error | |
| Default8BitActivationQuantizeConfig can only be used with `k | exception | error | |
| Activation {} not supported by Default8BitActivationQuantize | exception | error | |
| `set_quantize_weights` called on layer {} with {} weight par | exception | error | |
| Existing layer weight shape {} is incompatible withprovided | exception | error | |
| `set_quantize_activations` called on layer {} with {} activa | exception | error | |
| Variable name {} is not supported on CustomLayerQuantize({}) | exception | error | |
| {}/{} of Bottleneck weights is transformed. | exception | error | |
| Currently only supports FPN. | exception | error | |
| Currently only supports RetinaNetHead. | exception | error | |
| Attribute head type {} not supported. | exception | error | |
| Requested downsampling at a non-existing middle depthwise co | exception | error | |
| Requested downsampling at a non-existing starting depthwise | exception | error | |
| Unused logit_activation option inherited from vision Segment | exception | error | |
| The feature fusion method `pyramid_fusion` is not supported | exception | error | |
| `set_quantize_weights` called on layer {} with {} weight par | exception | error | |
| Existing layer weight shape {} is incompatible withprovided | exception | error | |
| `set_quantize_activations` called on layer {} with {} activa | exception | error | |
| DefaultNBitActivationQuantizeConfig can only be used with `k | exception | error | |
| Activation {} not supported by DefaultNBitActivationQuantize | exception | error | |
| Variable name {simple_name} is not supported on CustomLayerQ | exception | error | |
| {}/{} of Bottleneck weights is transformed. | exception | error | |
| `set_quantize_weights` called on layer {} with {} weight par | exception | error | |
| Existing layer weight shape {} is incompatible withprovided | exception | error | |
| `set_quantize_activations` called on layer {} with {} activa | exception | error | |
| Default8BitActivationQuantizeConfig can only be used with `k | exception | error | |
| Activation {} not supported by Default8BitActivationQuantize | exception | error | |
| Variable name {} is not supported. | exception | error | |
| Not enought original model weights. | exception | error | |
| Not enought quantized model weights. | exception | error | |
| {}/{} of Bottleneck weights is transformed. | exception | error | |
| Detection module not implemented for {} model. | exception | error | |
| Export module for {type(params.task)} is not supported. | exception | error | |
| The inner_dim of f{self.__class__} must be an even integer. | exception | error | |
| The type of input shape argument is not supported, got: %s | exception | error | |
| The input size (%d) is not a multiple of the number of atten | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| Unsupported conv_type: {} | exception | error | |
| Unrecognized final endpoint %s (available endpoints: %s). | exception | error | |
| Non local block is not implemented yet. | exception | error | |
| Attention cell is not implemented yet. | exception | error | |
| Attention cell super graph is not implemented yet. | exception | error | |
| Unknown impl {} for random brightness. | exception | error | |
| dtype {!r} is not supported! | exception | error | |
| The mode {} is not supported by the Parser. | exception | error | |
| Only 'backbone_projection' or 'backbone' can be used to init | exception | error | |
| The l2_weight_decay cannot be used together with lars optimi | exception | error | |
| Only 'all' or 'backbone' can be used to initialize the model | exception | error | |
| Skipping eval during pretraining without supervised head. | exception | error | |
| Unknown `output` value "%s". `output` can be either "logits" | exception | error | |
| At most one of `hub_module_url` and `init_checkpoint` can be | exception | error | |
| Unexpected inputs type to %s. | exception | error | |
| sequence_length must be specified for BigBird. | exception | error | |
| stride must be specified for BigBird. | exception | error | |
| sentencepiece_model_path must be specified for BigBird. | exception | error | |
| Static rank of `tensor` must be known. | exception | error | |
| `first_dim` out of bounds for `tensor` rank. | exception | error | |
| `last_dim` out of bounds for `tensor` rank. | exception | error | |
| `first_dim` must not be larger than `last_dim`. | exception | error | |
| `factor` must be positive. | exception | error | |
| `axis` out of bounds for `tensor` rank. | exception | error | |
| `max_distance` must not be negative. | exception | error | |
| `seq_len` must be positive. | exception | error | |
| `batch_size` must be positive. | exception | error | |
| `local_radius` must be positive. | exception | error | |
| `window_size` must be positive. | exception | error | |
| `window_size` must be odd. | exception | error | |
| Unknown evidence source: {info.source}. | exception | error | |
| Invalid encoding type. | exception | error | |
| Entity ID overflow: {entity_id}. Currently only entity_id<65 | exception | error | |
| Invalid augmentation_name: {} | exception | error | |
| Error trying to read input files for file pattern {file_patt | exception | error | |
| In TRAIN mode, padding/truncation is required. | exception | error | |
| No such detection unit: {self._detection_unit} | exception | error | |
| No such detection unit: {self._detection_unit}. Note that th | exception | error | |
| Not supported loss mode: {loss_mode} | exception | error | |
| Tensor rank must be at least %d. Got %d | exception | error | |
| `num_classes` should be given when requesting one hot label. | exception | error | |
| Invalid subset "{subset}". The available subsets are: {self. | exception | error | |
| only support num_test_clips = 1 for action localization task | exception | error | |
| If is_flow, frames should be given in float32. | exception | error | |
| min_resize should be larger than crop_size. Got ({min_resize | exception | error | |
| If `is_flow`, frames should be given in `tf.float32`. | exception | error | |
| max_resize should not be larger than pad_size. Got ({max_res | exception | error | |
| {sample_target_key} is not found in input dictionary. | exception | error | |
| {keyframe_index_key} is not found in input dictionary. | exception | error | |
| Unequal first dimension {}, {} | exception | error | |
| multi_crop is not supported with sync random states. | exception | error | |
| multi_crop should not be combined with ava augmentation. | exception | error | |
| sample_from_segments and sample_around_keyframes cannot be T | exception | error | |
| sample_from_segments is set to True while got num_test_clips | exception | error | |
| `parser_builder` has an unexpected type. | exception | error | |
| Unrecognized augmentation_type: %s | exception | error | |
| Choose to apply one of data augmentation policies: randaug a | exception | error | |
| num_classes should be given when requesting one hot label. | exception | error | |
| one_hot should be turned off if merge_multi_labels. | exception | error | |
| sample_random and sample_around_keyframe cannot be both True | exception | error | |
| Only support square crop. Got feature shape: %s | exception | error | |
| tf data service is not supported. | exception | error | |
| Invalid subset "{subset}". The available subsets are: {self. | exception | error | |
| Unrecognized ViT-3D implementation variant choice: %s | exception | error | |
| unrecognized pooler type: {pooler} | exception | error | |
| Expected features is a rank-5 tensor. Got shape %s | exception | error | |
| Should use the same batch normalization type. | exception | error | |
| %s classifier type not supported. | exception | error | |
| Unrecognized init_checkpoint_type: %s | exception | error | |
| ViT backbone is only partially loaded. | exception | error | |
| The number of classes from groundtruth labels and `num_class | exception | error | |
| The labels and logits must be at least rank 2. | exception | error | |
| Backbone depth should be equal to 3D UNet decoder's depth. | exception | error | |
| Exact one metric must be present, but {0} are present. | exception | error | |
| Bucket path must be non-empty starting with 'gs://' | exception | error | |
| Image not found at {image_path} | exception | error | |
| Bucket path must be non-empty starting with 'gs://' | exception | error | |
| Unsupported pooling strategy: {pooling!r}. Expected one of { | exception | error | |
| predict requires at least one image. | exception | error | |
| Checkpoint at '{checkpoint_path}' is missing 'model_state_di | exception | error | |
| Checkpoint state dict is missing 'head.weight'; cannot infer | exception | error | |
| Cannot infer pooling strategy: head input dim {head_input_fe | exception | error | |
| Bucket path must be non-empty starting with 'gs://' | exception | error | |
| Unknown augmentation name: {augmentation_name!r} | exception | error | |
| Classifier directory does not exist: {classifier_dir} | exception | error | |
| Train split folder is missing: {train_dir} | exception | error | |
| No class subfolders found under: {train_dir} | exception | error | |
| Augmented files already exist in the following class folders | exception | error | |
| {context} is missing required field(s): {', '.join(missing)} | exception | error | |
| {field_name} must be a number, got {value!r}. | exception | error | |
| {field_name} must be between {minimum} and {maximum}, got {v | exception | error | |
| {field_name} must be an integer, got {value!r}. | exception | error | |
| {field_name} must be at least 1, got {value!r}. | exception | error | |
| {field_name} must be a non-empty string. | exception | error | |
| {field_name} must be one of {list(allowed_values)}, got {val | exception | error | |
| {context} must be a list of exactly two integers, got {raw_c | exception | error | |
| crop_variants must be a non-empty list. Allowed values: {lis | exception | error | |
| Unknown crop variant {variant!r}. Allowed values: {list(ALLO | exception | error | |
| Duplicate crop variant {variant!r}. | exception | error | |
| {context} must be a list of exactly three integers in [{_MIN | exception | error | |
| {context}[{index}] must be an integer, got {channel_value!r} | exception | error | |
| {context}[{index}] must be in [{_MIN_RGB_VALUE}, {_MAX_RGB_V | exception | error | |
| {context} must be a non-empty list. Allowed values: {list(CA | exception | error | |
| Unknown augmentation {augmentation!r} in {context}. Allowed | exception | error | |
| Duplicate augmentation {augmentation!r} in {context}. | exception | error | |
| {context} must be a mapping. | exception | error | |
| prompts must be a non-empty mapping. | exception | error | |
| Prompt name must be a non-empty string, got {prompt_name!r}. | exception | error | |
| prompts.{prompt_name!r} must be a mapping. | exception | error | |
| cuda_visible_devices must be a string or integer, got {raw_v | exception | error | |
| cuda_visible_devices must be a string (quote it in YAML) or | exception | error | |
| Config file does not exist: {config_path} | exception | error | |
| Cannot read config file: {error} | exception | error | |
| Config file is not valid YAML: {error} | exception | error | |
| Config file must contain a top-level mapping of settings. | exception | error | |
| prompt_to_detect {prompt_to_detect!r} has no block under pro | exception | error | |
| The 'sam3' package is not installed or available in python p | exception | error | |
| Root directory does not exist: {root_dir} | exception | error | |
| No dataset subfolders found under: {root_dir} | exception | error | |
| Dataset {dataset_name!r} is missing required images folder: | exception | error | |
| Rejected directory already exists: {rejected_dir}. Remove or | exception | error | |
| crop_type must be one of {list(crop_type_index.keys())} | exception | error | |
| The 'sam3' package is not installed or available in python p | exception | error | |
| Root directory does not exist: {root_dir} | exception | error | |
| No dataset subfolders found under: {root_dir} | exception | error | |
| Dataset {dataset_name!r} is missing required input folder: { | exception | error | |
| Classifier output directory already exists: {classifier_outp | exception | error | |
| Unknown crop variant: {variant!r} | exception | error | |
| Dataset {dataset_name!r} is missing split folder: {split_inp | exception | error | |
| Root directory does not exist: {root_dir} | exception | error | |
| No dataset subfolders found under: {root_dir} | exception | error | |
| Dataset {dataset_name!r} is missing required input folder: { | exception | error | |
| Dataset {dataset_name!r} already has an output folder: {outp | exception | error | |
| Source folder not found: {source_folder} | exception | error | |
| Duplicate files found in '{destination_folder}':
{conflict | exception | error | |
| No images found in any folder for dataset {dataset_name!r}. | exception | error | |
| The masks must have the same dimensions. | exception | error | |
| No class subdirectories found in train_directory: {train_dir | exception | error | |
| Class directory is empty: {class_path} | exception | error | |
| image_size must be a positive multiple of {patch_size}, got | exception | error | |
| model_name must be a non-empty string. | exception | error | |
| repo_dir must be a non-empty path. | exception | error | |
| Unsupported pooling strategy: {pooling!r}. Expected one of { | exception | error | |
| Input directory not found at '{INPUT_DIR}' | exception | error | |
| Error executing command: {cmd}. Error: {stderr} | exception | error | |
| The masks must have the same dimensions. | exception | error | |
| Class '{class_name}' is not assigned to any collapsed catego | exception | error | |
| collapsed_categories.enable is true but 'mapping' is empty o | exception | error | |
| Category '{category_name}' must map to a list of class names | exception | error | |
| Category '{category_name}' must map to a non-empty list of c | exception | error | |
| Class '{class_name}' in category '{category_name}' is not pr | exception | error | |
| Class '{class_name}' is assigned to both '{seen_classes[clas | exception | error | |
| The following classes are not assigned to any collapsed cate | exception | error | |
| Config file not found or inaccessible: {yaml_path} | exception | error | |
| Invalid YAML syntax in {yaml_path}: {err} | exception | error | |
| Configuration root in {yaml_path} must be a dictionary. | exception | error | |
| Error validating configuration structure in {yaml_path}: {er | exception | error | |
| Cannot infer pooling strategy. Head input dimension {head_in | exception | error | |
| Checkpoint is missing required key 'model_state_dict'. | exception | error | |
| Checkpoint state dict is missing required key 'head.weight'. | exception | error | |
| Failed to load state dict into model: {error} | exception | error | |
| OpenCV could not read the image at {image_path!s}. The file | exception | error | |
| ensure_directories_exist received an empty path or the curre | exception | error | |
| The 'sam3' package is not installed or available in python p | exception | error | |
| If masks are included you cannot convert coco from the91 cla | exception | error | |
| Decoder version cannot be None, specify v3 or v4. | exception | error | |
| Unsupported model version please select from {v3, v4}, or sp | exception | error | |
| Unsupported model type please select from {yolo_model.YOLO_M | exception | error | |
| Model has to be built before number of boxes can be determin | exception | error | |
| the input to the CSPStack must be a list of layers that we c | exception | error | |
| More than one maxpool should be specified in SPP block | exception | error | |
| maximum change in aspect ratio must be between 0 and 0.5 | exception | error | |
| Smart Bias is only supported currently with alinear warm up. | exception | error | |
| `momentum` must be between [0, 1]. | exception | error | |
| Export module not implemented for {} task. | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| Wrong train dataset size {!r} | exception | error | |
| Wrong validation dataset size {!r} | exception | error | |
| Unknown feature source {self._feature_sources[i]} for {name} | exception | error | |
| Invalid label field: {self._label_field}! | exception | error | |
| top_n must be a positive integer or None. | exception | error | |
| the shape of predictions and actuals does not match. | exception | error | |
| 'num_positives' was provided but it was a negative number. | exception | error | |
| n must be 'None' or a positive integer. It was '%s'. | exception | error | |
| k must be a positive integer. | exception | error | |
| total_sample must be positive. | exception | error | |
| num_class must be a positive integer. | exception | error | |
| Unrecognized pooling method: %s | exception | error | |
| epochs_between_evals > 1 not supported for file based datase | exception | error | |
| producer batch size ({}) differs from params batch size ({}) | exception | error | |
| User positives ({}) is different from item positives ({}) | exception | error | |
| Eval batch size {} is not divisible by {} | exception | error | |
| Fatal exception in the data production loop: {} | exception | error | |
| Unrecognized constructor: {} | exception | error | |
| Expected to find {} users, but found {} | exception | error | |
| Expected to find {} items, but found {} | exception | error | |
| dataset {} is not in {{{}}} | exception | error | |
| Pre-batch ({}) size is not equal to batch size ({}) | exception | error | |
| Evaluation batch size must be divisible by {} times {} | exception | error | |
| TPU training does not support data producer yet. Use pre-pro | exception | error | |
| The first layer size should be multiple of 2! | exception | error | |
| length of vocab_sizes: {len(vocab_sizes)} is not equal to th | exception | error | |
| embedding_dim is not either a list or an int, got {type(embe | exception | error | |
| length of vocab_sizes: {len(vocab_sizes)} is not equal to th | exception | error | |
| max_ids_per_table is not either a list or an int or None, go | exception | error | |
| length of vocab_sizes: {len(vocab_sizes)} is not equal to th | exception | error | |
| max_unique_ids_per_table is not either a list or an int or N | exception | error | |
| {self.task_config.model.interaction} is not supported it mu | exception | error | |
| The mode is not implemented: %s | exception | error | |
| feature_names must be a non-empty list of strings but got {f | exception | error | |
| feature_names must be a list of strings, but got types {list | exception | error | |
| Layer inputs is missing features: {missing_features} | exception | error | |
| Got unsupported tensor shape type for feature {feature_name} | exception | error | |
| All features from the feature_names set must be tensors with | exception | error | |
| Got unsupported tensor type for feature {feature_name}. The | exception | error | |
| The linear layering config cannot be `None` when using the l | exception | error | |
| The control logits and treatment logits computed by the cont | exception | error | |
| The {name}_input_combiner layer must be specified if the {na | exception | error | |
| The {name}_feature_encoder layer must be specified if the {n | exception | error | |
| y_pred must be of type `TwoTowerTrainingOutputs` but got typ | exception | error | |
| y_pred must be of type `TwoTowerTrainingOutputs` but got typ | exception | error | |
| `loss_fn` cannot be a Keras `Loss` object, pass a non-reduci | exception | error | |
| Value passed to `from_logits` ({from_logits}) is conflicting | exception | error | |
| `slice_by_treatment` must be False when y_pred is a `tf.Tens | exception | error | |
| y_pred must be of type `TwoTowerTrainingOutputs`, `tf.Tensor | exception | error | |
| `from_logits` must be set to `True` when `y_pred` is of type | exception | error | |
| Full loss computation is not yet supported. | exception | error | |
| `slice_by_treatment` must be set to `False` when `y_pred` is | exception | error | |
| The slicing spec must be a non-empty dictionary. | exception | error | |
| All slicing values in the slicing spec must be one of `int`, | exception | error | |
| The slicing values passed to the slicing spec must be unique | exception | error | |
| The `slicing_feature` and slicing values in `slicing_spec` m | exception | error | |
| The output of the given metric must either be a `tf.Tensor` | exception | error | |
| y_pred must be of type `TwoTowerTrainingOutputs` but got typ | exception | error | |
| y_pred must be of type `TwoTowerTrainingOutputs` but got typ | exception | error | |
| The treatment_indicator feature (specified as '{self._treatm | exception | error | |
| is_treatment tensor must be a tensor of shape (D0,) (D0, 1) | exception | error | |
| values and is_treatment must be tensors of shapes (D0, D1, . | exception | error | |
| is_treatment must be a tensor castable to boolean but got te | exception | error | |
| Threshold for checking stop conditions must be a number. | exception | error | |
| Eval metric being checked against stop conditions must be a | exception | error | |
| `maskrcnn.Parser` only supports `pad = True`. | exception | error | |
| `maskrcnn.Parser` only supports `keep_aspect_ratio = True`. | exception | error | |
| Wrong train dataset size {!r} | exception | error | |
| Wrong validation dataset size {!r} | exception | error | |
| {} does not exist across image directories. | exception | error | |
| {} is not unique across image directories | exception | error | |
| Image format is invalid: {format_str} | exception | error | |
| JSON files for {} shots (seed {}) have different info, licen | exception | error | |
| For raw image feature, height, width and num_channels fields | exception | error | |
| Cannot convert type {} to feature | exception | error | |
| Unknown value_type parameter - {} | exception | error | |
| dtype {!r} is not supported! | exception | error | |
| Augmentation policy {} not supported. | exception | error | |
| input_path and weights must both be a Config to use weighted | exception | error | |
| The number of input_path and weights must be the same, but g | exception | error | |
| input_path key '%s' does not have a corresponding weight. | exception | error | |
| Invalid total_batch_size: {} | exception | error | |
| The batch size of pseudo-label dataset should not be larger | exception | error | |
| Must use drop_remainder=True with CombinationDatasetInputRea | exception | error | |
| Invalid batch_size: {} and pseudo_label_data_ratio: {}, resu | exception | error | |
| Augmentation policy {} not supported. | exception | error | |
| Augmentation policy {aug_type.type} not supported. | exception | error | |
| groundtruth_padded_size ([height, width]) needs to bespecifi | exception | error | |
| centered_crop is only supported when resize_eval_groundtruth | exception | error | |
| The label map file is in incorrect format. | exception | error | |
| Each row of the csv label map file must be in `id,name` form | exception | error | |
| TFDS Classification {tfds_name} is not supported | exception | error | |
| TFDS Detection {tfds_name} is not supported | exception | error | |
| TFDS Segmentation {tfds_name} is not supported | exception | error | |
| Random stride range should be >= 0, got {} | exception | error | |
| Unknown input image format: {} | exception | error | |
| `num_classes` should be given when requesting one hot label. | exception | error | |
| Augmentation policy {} not supported. | exception | error | |
| max_num_eval_detections must be an integer. | exception | error | |
| Missing the required key `{}` in predictions! | exception | error | |
| Missing the required key `{}` in groundtruths! | exception | error | |
| One and only one of `annotation_file` and `gt_dataset` needs | exception | error | |
| The `eval_type` can only be either `box` or `mask`. | exception | error | |
| Results do not correspond to the current dataset! | exception | error | |
| Missing the required key `{}` in predictions! | exception | error | |
| Missing the required key `{}` in groundtruths! | exception | error | |
| Length of class_weights should be {} | exception | error | |
| Groundtruth mask must have only 1 layer if using categorical | exception | error | |
| Groundtruth matting map only supports 2 classes, but got {} | exception | error | |
| Groundtruth matting map must have only 1 layer, but got {} l | exception | error | |
| Length of class_weights should be {} | exception | error | |
| Block func {} not supported. | exception | error | |
| The block spec cannot be empty for {} ! | exception | error | |
| The block spec values {} do not match with the schema {} | exception | error | |
| The MobileDet version {} is not supported | exception | error | |
| filter_size_scale is not greater than zero. | exception | error | |
| Expected rank 4 input, was: %d | exception | error | |
| Unknown block type {} for layer {} | exception | error | |
| The block spec cannot be empty for {} ! | exception | error | |
| The block spec values {} do not match with the schema {} | exception | error | |
| The MobileNet version {} is not supported | exception | error | |
| filter_size_scale is not greater than zero. | exception | error | |
| Only allowed output_stride values are 8, 16, 32. | exception | error | |
| Output stride must be None, 1 or a multiple of 2. | exception | error | |
| Expected rank 4 input, was: %d | exception | error | |
| Unknown block type {} for layer {} | exception | error | |
| Block fn `{}` is not supported. | exception | error | |
| Stem type {} not supported. | exception | error | |
| Number of blocks in temporal specs should equal to resnet_sp | exception | error | |
| Block fn `{}` is not supported. | exception | error | |
| Stem type {stem_type} not supported. | exception | error | |
| Number of elements in `temporal_kernel_sizes` must equal to | exception | error | |
| Stem type {} not supported. | exception | error | |
| Block fn `{}` is not supported. | exception | error | |
| multigrid has to match number of block_repeats | exception | error | |
| Stem type {} not supported. | exception | error | |
| Block fn `{}` is not supported. | exception | error | |
| Block fn `{}` is not supported. | exception | error | |
| Number of output filters must be even to ensure splitting in | exception | error | |
| Activation {} not implemented. | exception | error | |
| Duplicate feats found for output level {}. | exception | error | |
| SpineNet-{} is not a valid architecture. | exception | error | |
| Duplicate feats found for output level {}. | exception | error | |
| Mobile SpineNet-{} is not a valid architecture. | exception | error | |
| unrecognized pooler type: {pooler} | exception | error | |
| Inconsistent decoder type {decoder_type}. Need to be `aspp`. | exception | error | |
| Fusion type {} not supported. | exception | error | |
| Backbone min level should be less or equal to FPN min level | exception | error | |
| Inconsistent decoder type {decoder_type}. Need to be `fpn`. | exception | error | |
| Backbone min level should be less or equal to FPN min level | exception | error | |
| input_offset ({}) is larger than num feats({}) | exception | error | |
| unknown combine_fn `{}`. | exception | error | |
| Inconsistent decoder type {decoder_type}. Need to be `nasfpn | exception | error | |
| Attribute head type {} not supported. | exception | error | |
| share_classification_heads cannot be set as True when att_pr | exception | error | |
| Invalid `num_convs` {att_num_convs} for {att_name}. | exception | error | |
| Invalid `num_filters` {att_num_filters} for {att_name}. | exception | error | |
| Only support dictionary decoder_output. | exception | error | |
| Boxes should have either 1 class or same as scores. | exception | error | |
| Number of locations is different. | exception | error | |
| Number of sides is incorrect. | exception | error | |
| The last dimension of predicted boxes should be divisible by | exception | error | |
| The last dimension of predicted scores should be divisible b | exception | error | |
| NMS version {} not supported. | exception | error | |
| If not using TFLite custom NMS, `use_class_agnostic_nms` can | exception | error | |
| Attribute learning is only supported for NMSv1 but NMS {} is | exception | error | |
| Inconsistent shapes in shard_tensors: first is {tensors[0].s | exception | error | |
| Boxes shape ({boxes.shape}) last dimension must be 4 to repr | exception | error | |
| Boxes shape ({boxes.shape}) and scores shape ({scores.shape} | exception | error | |
| Requested downsampling at a non-existing middle depthwise. | exception | error | |
| Requested downsampling at a non-existing starting depthwise. | exception | error | |
| Unexpected inputs to %s with length at %d | exception | error | |
| In the case of `norm_first`, the residual connection should | exception | error | |
| Padding for kernel size {} not known. | exception | error | |
| Initial drop rate must be within 0 and 1. | exception | error | |
| target_level should be less than max_level | exception | error | |
| "channels_first" mode is unsupported. | exception | error | |
| Input should have rank {}, got {} | exception | error | |
| The number of filters must be evenly divisible by the number | exception | error | |
| The argument `kernel_size` cannot contain 0(s). Received: %s | exception | error | |
| `class_agnostic_bbox_pred` needs to be True if multiple dete | exception | error | |
| `outer_boxes_scale` should be a value >= 1.0. | exception | error | |
| `mask_sampler` is not provided in Mask R-CNN. | exception | error | |
| `mask_roi_aligner` is not provided in Mask R-CNN. | exception | error | |
| Input should be a tf.Tensor or a sequence of tf.Tensor, not | exception | error | |
| the structure of `anchor_sizes` must be a tuple, list, or di | exception | error | |
| padding should be one of "REFLECT", "CONSTANT", or "SYMMETRI | exception | error | |
| The {name} argument must be a tuple of {n} integers. Receive | exception | error | |
| The {name} argument must be a tuple of {n} integers. Receive | exception | error | |
| sigma should be a float or a tuple/list of 2 floats | exception | error | |
| sigma should be greater than or equal to 0. | exception | error | |
| translations rank must be statically known | exception | error | |
| translations should have rank 1 or 2. | exception | error | |
| Angles should have a rank 0 or 1. | exception | error | |
| output_shape must be a 1-D Tensor of 2 elements: new_height, | exception | error | |
| Bad image rank: {} | exception | error | |
| Image rank 4 is not supported | exception | error | |
| translate_y_only_bboxes does not support rank 4 boxes | exception | error | |
| Invalid augmentation_name: {} | exception | error | |
| Wrong shape detected for custom policy. Expected (:, :, 3) b | exception | error | |
| `threshold` must be sorted, got {} | exception | error | |
| len(`indicators`) must be len(`thresholds`) + 1, got indicat | exception | error | |
| boxes.shape[-1] is {:d}, but must be 4. | exception | error | |
| Last dimension of boxes must be 4 but is {:d} | exception | error | |
| scale is {}, but outer box scale must be greater than 1.0. | exception | error | |
| encoded_boxes.shape[-1] is {:d}, but must be 4. | exception | error | |
| boxes.shape[1] is {:d}, but must be 4. | exception | error | |
| Expected the last dimension of `bbox` has size == 4, but the | exception | error | |
| `groudtruth_boxes` must be rank 2 or 3, got {} | exception | error | |
| `anchors` must be rank 2 or 3, got {} | exception | error | |
| `groundtruth_boxes` is unbatched while `anchors` is batched | exception | error | |
| Only 1-crop and 3-crop are supported. Found {num_crops!r}. | exception | error | |
| positive_fraction should be in range [0,1]. Received: %s. | exception | error | |
| indicator must be static in shape when is_static isTrue | exception | error | |
| labels must be static in shape when is_static isTrue | exception | error | |
| batch_size has to be an integer when is_static isTrue. | exception | error | |
| indicator must be 1 dimensional, got a tensor of shape %s | exception | error | |
| labels must be 1 dimensional, got a tensor of shape %s | exception | error | |
| labels should be of type bool. Received: %s | exception | error | |
| indicator should be of type bool. Received: %s | exception | error | |
| Expected the input_matrix tensor (input_h, input_w) has rank | exception | error | |
| Expected the row_indices tensor (output_h) has rank == 1, wa | exception | error | |
| Expected the input images (batch_size, height, width, ...) h | exception | error | |
| Expected the last dimension of `bbox` has size == 4, but the | exception | error | |
| `TargetGather` does not support `labels` with rank larger th | exception | error | |
| Detection module not implemented for {} model. | exception | error | |
| Unrecognized `input_type` | exception | error | |
| Export module not implemented for {} task. | exception | error | |
| Export module not implemented for {} task. | exception | error | |
| add_tpu_function_alias is only allowed for input_type of: im | exception | error | |
| Export module not implemented for {} task. | exception | error | |
| Task {} not supported. | exception | error | |
| `saved_model_dir`, `model` or `concrete_function` must be sp | exception | error | |
| quantization type {quant_type} is not supported. | exception | error | |
| Unrecognized `input_type` | exception | error | |
| Batch size cannot be more than 1. | exception | error | |
| Video classification do not support image bytes input. | exception | error | |
| Unrecognized `input_type` | exception | error | |
| Only 'all' or 'backbone' can be used to initialize the model | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| No instance metrics defined when use_masks is %s | exception | error | |
| Unknown decoder type: {}! | exception | error | |
| Attribute {head.name} not found in label targets. | exception | error | |
| Attribute {head.name} not found in model outputs. | exception | error | |
| Attribute type {head.type} not supported. | exception | error | |
| Can't evaluate using annotation file when TFDS is used. | exception | error | |
| Only 'all' or 'backbone' can be used to initialize the model | exception | error | |
| Unknown input file type {!r} | exception | error | |
| The number of logical devices %d is not supported. Supported | exception | error | |
| Spatial partitioning is only supported for TPUStrategy. | exception | error | |
| Train and eval input partition dims can not bepartitioned on | exception | error | |
| Need to also define matched_threshold whenunmatched_threshol | exception | error | |
| unmatched_threshold needs to be smaller or equalto matched_t | exception | error | |
| When negatives are in between matched and unmatched threshol | exception | error | |
| positive_fraction should be in range [0,1]. Received: %s. | exception | error | |
| indicator must be static in shape when is_static isTrue | exception | error | |
| labels must be static in shape when is_static isTrue | exception | error | |
| batch_size has to be an integer when is_static isTrue. | exception | error | |
| indicator must be 1 dimensional, got a tensor of shape %s | exception | error | |
| labels must be 1 dimensional, got a tensor of shape %s | exception | error | |
| labels should be of type bool. Received: %s | exception | error | |
| indicator should be of type bool. Received: %s | exception | error | |
| The number of anchors inferred from encoded_boxes and anchor | exception | error | |
| Invalid dimensions for box data. | exception | error | |
| Invalid tensor type: should be tf.float32 | exception | error | |
| field %s does not exist | exception | error | |
| Scores should have rank 1 or 2 | exception | error | |
| Scores should have rank 1 or have shape consistent with [Non | exception | error | |
| thresh must be between 0 and 1 | exception | error | |
| boxlist must be a BoxList | exception | error | |
| input boxlist must have 'scores' field | exception | error | |
| nms_iou_thresh must be between 0 and 1 | exception | error | |
| voting_iou_thresh must be between 0 and 1 | exception | error | |
| pool_boxes must be a BoxList | exception | error | |
| pool_boxes must have a 'scores' field | exception | error | |
| pool_boxes must have a 'classes' field | exception | error | |
| iou_thresh must be between 0 and 1 | exception | error | |
| selected_boxes must be a BoxList | exception | error | |
| match_results should have rank 1 | exception | error | |
| match_results should be an int32 or int64 scalar tensor | exception | error | |
| keypoints are provided but keypoints_flip_permutation is not | exception | error | |
| Image should be 3D tensor | exception | error | |
| Unequal shapes {}, {} | exception | error | |
| anchors must be an BoxList | exception | error | |
| groundtruth_boxes must be an BoxList | exception | error | |
| Input must be of size [N, 4] | exception | error | |
| `image_mean` and `image_std` should be the same type. | exception | error | |
| Both `image_mean` and `image_std` should be set or None at t | exception | error | |
| `image` not of type np.uint8 | exception | error | |
| `mask` not of type np.uint8 | exception | error | |
| `mask` elements should be in [0, 1] | exception | error | |
| The image has spatial dimensions %s but the mask has dimensi | exception | error | |
| `trainer` and `evaluator` should not both be `None`. | exception | error | |
| `steps_per_loop` is required when `trainer` is provided. | exception | error | |
| `steps_per_loop` ({steps_per_loop}) must be a positive integ | exception | error | |
| `summary_interval` ({summary_interval}) must be larger than | exception | error | |
| `summary interval` ({summary_interval}) must be a multiple o | exception | error | |
| `global_step` must be a `tf.Variable`. | exception | error | |
| `steps` ({steps}) should be > 0, or == -1. | exception | error | |
| `{attribute}` is not set. Pass `{attribute}` to `Controller. | exception | error | |
| `use_tf_while_loop=True` and `use_tf_function=False` is not | exception | error | |
| `use_tpu_summary_optimization=True` and `use_tf_while_loop=F | exception | error | |
| Looping until exhausted is not supported if `options.use_tf_ | exception | error | |
| `dataset_or_fn` should be either callable or an instance of | exception | error | |
| `epoch_end` can only be called inside an epoch. | exception | error | |
| `num_steps` should be a `tf.Tensor`. Passing a Python value | exception | error | |
| Encountered error when importing tensorflow_text: %s | console | error | |
| Resetting the generator. %s: %s | console | error | |
| You are reloading a model that was saved with a potentially- | console | error | |
| norm_activation is not used in MoViNets, but specified: %s | console | error | |
| norm_activation is ignored. | console | error | |
| You are reloading a model that was saved with a potentially- | console | error | |
| You are reloading a model that was saved with a potentially- | console | error | |
| Skip build head type: %s | console | error | |