AVhdZddlmZddlmZed ddZed ddZed dd Zed  dd Zed dd Z eddZ ed ddZ y)zDmlir is an experimental library that provides support APIs for MLIR.) pywrap_mlir) tf_exportz#mlir.experimental.convert_graph_defc0tj|||S)aKImport a GraphDef and convert it to a textual MLIR module. This API is only intended for inspecting the internals of TensorFlow and the string returned is at the moment intended for debugging purposes. Args: graph_def: An object of type graph_pb2.GraphDef or a textual proto representation of a valid GraphDef. pass_pipeline: A textual description of an MLIR Pass Pipeline to run on the module, see MLIR documentation for the [textual pass pipeline syntax](https://mlir.llvm.org/docs/PassManagement/#textual-pass-pipeline-specification). show_debug_info: Whether to include locations in the emitted textual form. Returns: A textual representation of the MLIR module corresponding to the graphdef. Raises: InvalidArgumentError: if graph_def is invalid or cannot be converted to MLIR. )rimport_graphdef) graph_def pass_pipelineshow_debug_infos T/home/dcms/DCMS/lib/python3.12/site-packages/tensorflow/python/compiler/mlir/mlir.pyconvert_graph_defr s0  $ $Y  OOz"mlir.experimental.convert_functionc0tj|||S)aImport a ConcreteFunction and convert it to a textual MLIR module. This API is only intended for inspecting the internals of TensorFlow and the string returned is at the moment intended for debugging purposes. A [tf.function](https://www.tensorflow.org/api_docs/python/tf/function) can be imported and converted from TensorFlow to TensorFlow MLIR with this API by extracting its ConcreteFunction (eagerly-executing wrapper around a [tf.Graph](https://www.tensorflow.org/api_docs/python/tf/Graph)). For example: >>> @tf.function ... def add(a, b): ... return a + b >>> concrete_function = add.get_concrete_function( ... tf.TensorSpec(None, tf.dtypes.float32), ... tf.TensorSpec(None, tf.dtypes.float32)) >>> tf.mlir.experimental.convert_function(concrete_function) '...module attributes {...} {...}...' Args: concrete_function: An object of type ConcreteFunction. pass_pipeline: A textual description of an MLIR Pass Pipeline to run on the module, see MLIR documentation for the [textual pass pipeline syntax](https://mlir.llvm.org/docs/PassManagement/#textual-pass-pipeline-specification). show_debug_info: Whether to include locations in the emitted textual form. Returns: A textual representation of the MLIR module corresponding to the ConcreteFunction. Raises: InvalidArgumentError: if concrete_function is invalid or cannot be converted to MLIR. )rimport_function)concrete_functionrr s r convert_functionr0sT  $ $ r z%mlir.experimental.convert_saved_modelc0tj|||S)a0Converts a SavedModel to MLIR module. Args: saved_model_path: Path to SavedModel. exported_names: Names to export. show_debug_info: Whether to include locations in the emitted textual form. Returns: A textual representation of the MLIR module corresponding to the SavedModel. )r(experimental_convert_saved_model_to_mlir)saved_model_pathexported_namesr s r convert_saved_modelr_s  = = r z(mlir.experimental.convert_saved_model_v1c 8tj|||||||S)axConverts a v1 SavedModel to MLIR module. Args: saved_model_path: Path to SavedModel. exported_names: Names to export. tags: MetaGraphDef to be loaded is identified by the supplied tags. lift_variables: Whether to promote tf.VarHandleOp to resource arguments. include_variables_in_initializers: Keeps the variables in initializers before lifting variables. upgrade_legacy: Functionalize the input graph before importing. show_debug_info: Whether to include locations in the emitted textual form. Returns: A textual representation of the MLIR module corresponding to the SavedModule. )r+experimental_convert_saved_model_v1_to_mlir)rrtagslift_variables!include_variables_in_initializersupgrade_legacyr s r convert_saved_model_v1rss,4  @ @ ' r z#mlir.experimental.run_pass_pipelinec0tj|||S)aRRuns a pipeline over input module. Args: mlir_txt: Textual representation of the MLIR module. pass_pipeline: Pass pipeline to run on module. show_debug_info: Whether to include locations in the emitted textual form. Returns: A textual representation of the MLIR module corresponding to the transformed module. )rexperimental_run_pass_pipeline)mlir_txtrr s r run_pass_pipeliner s  3 3  r z mlir.experimental.write_bytecodec0tj||y)zWrites an MLIR module out as bytecode. Args: filename: The filename to write to. mlir_txt: The MLIR module in textual format. N)rexperimental_write_bytecode)filenamers r r"r"s))(H=r z)mlir.experimental.tflite_to_tosa_bytecodeNc6tj|||||y)aConverts TFLite flatbuffer to TOSA dialect in MLIR bytecode. Args: flatbuffer: Path to flatbuffer. bytecode: Path to output bytecode. use_external_constant: Whether to create `tfl.external_const` instead of `tfl.const`. ordered_input_arrays: ordered_output_arrays: If ordered_output_arrays is not empty, then the function will only return nodes in ordered_output_arrays in the same order N)r$experimental_tflite_to_tosa_bytecode) flatbufferbytecodeuse_external_constantordered_input_arraysordered_output_arrayss r tflite_to_tosa_bytecoder+s!&22 r )ztf-standard-pipelineF)F)TF)FNN) __doc__tensorflow.pythonr tensorflow.python.util.tf_exportrr rrrr r"r+r r r0sK)6 01EJP2P4 /0)+1+\ 236;4& 56!7!H 012" -.>/> 67  8r