- Sort Score
- Result 10 results
- Languages All
Results 31 - 40 of 89 for legalization (0.42 sec)
-
tensorflow/compiler/mlir/tf2xla/transforms/legalize_tf_with_tf2xla_passes.h
/// `prefer_tf2xla` means an op will be included iff it is not in /// `MlirLegalizedUnderPreferTf2XlaSet`. `!prefer_tf2xla` mean an op will be /// included if there is no native MLIR legalization for the op. void PopulateLegalizeTfWithTf2XlaPatterns(llvm::StringRef device_type, RewritePatternSet& patterns, MLIRContext* ctx,
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Tue May 28 21:49:50 UTC 2024 - 2.3K bytes - Viewed (0) -
tensorflow/compiler/mlir/tf2xla/transforms/tf2xla_rewriter.h
// Tries to legalize the specified TensorFlow op, if supported. // // Emits an error and returns failure if an error is encountered during // conversion. Note that success return value doesn't mean successful // legalization. mlir::LogicalResult LegalizeOp(); // Converts the given operand to expression of kind kConstant or kXlaOp. // Emits a remark and returns expression of kind kInvalid on failure.
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu May 02 09:16:07 UTC 2024 - 5K bytes - Viewed (0) -
tensorflow/compiler/mlir/quantization/stablehlo/passes/bridge/passes.h
// convert TF uniform quantized ops to the corresponding quantized MHLO ops. std::unique_ptr<OperationPass<func::FuncOp>> CreateConvertTFQuantOpsToMHLOPass(); // TODO(b/288094093): Migrate uniform quantization legalization in a separate // pass. void PopulateLegalizeTfQuantizationPatterns(MLIRContext *context, RewritePatternSet *patterns);
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Fri Feb 23 01:41:18 UTC 2024 - 2.6K bytes - Viewed (0) -
tensorflow/compiler/mlir/tf2xla/internal/legalize_tf_mlir.h
namespace internal { // Compiles a serialized MLIR module and returns a serialized MLIR module of the // result of running all the MLIR Bridge passes. If compile_to_xla_hlo is true // then those passes include all the Legalization to XLA HLO which is returned // in the compilation_result. absl::StatusOr<std::string> CompileFromMlirToXlaHlo( bool lower_to_xla_hlo, const tpu::MlirToHloArgs& computation,
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Sun Apr 14 20:29:34 UTC 2024 - 2.8K bytes - Viewed (0) -
tensorflow/compiler/mlir/tf2xla/transforms/xla_legalize_tf_test.cc
EXPECT_TRUE(status); EXPECT_EQ(legalize_failure_count.Read("tf.DoesntExist", "Unknown"), 1); } TEST(XlaLegalizeTest, LegalOp) { // We expect legalization to fail for legal op with dynamic shapes: static constexpr char kMlirLegalOpStr[] = R"( func.func @infeed_dequeue_tuple_dynamic_error() -> (tensor<3x3xf32>, tensor<4x?xf32>) {
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu May 02 09:16:07 UTC 2024 - 4.2K bytes - Viewed (0) -
tensorflow/compiler/mlir/tf2xla/tests/legalize-tf.mlir
// This test runs twice: // 1. Through FILECHECK_OPTS="" FileCheck with chlo legalization disabled since verifying // that the chlo ops emit produces more useful tests. // 2. With chlo legalization enabled, verifying diagnostics to pick up any // issues with the full lowering (can catch some broadcasting corner // cases which emit with a warning).
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Mon May 06 18:46:23 UTC 2024 - 335.5K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/stablehlo/transforms/tflite_legalize_hlo.cc
if (failed(applyPartialConversion(getOperation(), target, std::move(patterns)))) { getOperation().emitError("mhlo to TFLite legalization failed."); signalPassFailure(); } } } // namespace void PopulateLegalizeHloToTFLitePatterns(RewritePatternSet* patterns, MLIRContext* context) {
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 25 16:01:03 UTC 2024 - 6.2K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/transforms/prepare_tf.cc
// is not supported in TensorFlow because TensorFlow requires filters in the // HWIO data format. // // Motivation to prepare for the TFLite legalization before the actual // legalization is to exploit constant folding opportunities in any newly // created ops by leveraging constant folding support for the TensorFlow ops. // This way TFLite can be used as a serialization format only and does not
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Tue May 28 21:49:50 UTC 2024 - 64.6K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/transforms/analyze_variables.cc
if (IsSupportedTFDataForwardingOp(op)) { return WalkResult::advance(); } // If any of the operands is a resource type, then we break // and mark the module as not valid for TFLite legalization. // Note: this might disable native variables in more than needed cases. // TODO(b/189370197): Enhance variable analysis. for (auto operand : op->getOperands()) { if (mlir::isa<TF::ResourceType>(
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 25 16:01:03 UTC 2024 - 4.3K bytes - Viewed (0) -
tensorflow/compiler/mlir/quantization/stablehlo/passes/bridge/convert_tf_quant_ops_to_mhlo.cc
// Currently for dot, PTQ supports per-tensor quantization. if (rhs_quantized_dimension != -1) { return rewriter.notifyMatchFailure( op, "Legalization supports only rhs_quantization_axis -1."); } auto rhs_type = GetUniformQuantizedType( op, op.getRhs().getType(), op.getRhsScales(), op.getRhsZeroPoints(),
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Fri May 17 17:58:54 UTC 2024 - 30.9K bytes - Viewed (0)