- Sort Score
- Result 10 results
- Languages All
Results 51 - 56 of 56 for PartitionedCall (0.37 sec)
-
tensorflow/compiler/mlir/tensorflow/transforms/shape_inference.cc
// `potential_refined_type`. Return true if the type was changed. bool RefineResultType(Operation* op, Value result, Type potential_refined_type); // Infers the shape from a (Stateful)PartitionedCall operation by looking up // the called function and propagating the return type. bool InferShapeForCall(CallOpInterface call_op); bool InferShapeForCast(Operation* op);
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Sat Jun 08 07:28:49 UTC 2024 - 134.1K bytes - Viewed (0) -
tensorflow/compiler/mlir/tensorflow/ir/tf_ops.td
The LegacyCall operation represents a direct call to a function that is within the same symbol scope as the call and is mapped to a GraphDef node with the function name as the op name. Unlike a PartitionedCall which represents asynchronously executing a function across multiple devices, a LegacyCall ignores specification for ops in the attached function and instead executes it on the device assigned to this op.
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Wed Apr 24 04:08:35 UTC 2024 - 90.5K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/flatbuffer_export.cc
} } return false; } // Returns whether the current op is not supported by the TF Lite runtime. static bool IsUnsupportedFlexOp(const std::string& op_name) { return op_name == "PartitionedCall" || op_name == "StatefulPartitionedCall"; } // Create description of operation that could not be converted. static std::string GetOpDescriptionForDebug(Operation* inst) { const int kLargeElementsAttr = 16;
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Wed Jun 12 21:41:49 UTC 2024 - 164.5K bytes - Viewed (0) -
tensorflow/compiler/mlir/quantization/tensorflow/python/integration_test/quantize_model_test.py
# true branch doesn't. def _is_quantized_function_call_node( node_def: node_def_pb2.NodeDef, ) -> bool: return node_def.op == 'PartitionedCall' and node_def.attr[ 'f' ].func.name.startswith('quantized_') for func in output_graphdef.library.function: if func.signature.name.startswith('cond_false'):
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Fri May 17 03:36:50 UTC 2024 - 235.6K bytes - Viewed (0) -
tensorflow/compiler/mlir/tf2xla/transforms/legalize_tf.cc
Value return_values[] = {selected_input, selected_index}; b.create<ReturnOp>(return_values); } //===----------------------------------------------------------------------===// // PartitionedCall op utilities. //===----------------------------------------------------------------------===// // Verify that the arguments to be passed into the function are the same types // as the function paramter types.
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Tue Jun 11 20:00:43 UTC 2024 - 291.8K bytes - Viewed (0) -
tensorflow/compiler/mlir/tensorflow/ir/tf_generated_ops.td
); TF_DerivedOperandTypeAttr T = TF_DerivedOperandTypeAttr<0>; TF_DerivedOperandTypeAttr dtype = TF_DerivedOperandTypeAttr<1>; } def TF_PartitionedCallOp : TF_Op<"PartitionedCall", [CallOpInterface, DeclareOpInterfaceMethods<SymbolUserOpInterface>, Pure]> { let summary = [{ returns `f(inputs)`, where `f`'s body is placed and partitioned. }]; let description = [{
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Tue Jun 11 23:24:08 UTC 2024 - 793K bytes - Viewed (0)