- Sort Score
- Result 10 results
- Languages All
Results 41 - 50 of 78 for tpu0 (0.04 sec)
-
tensorflow/compiler/mlir/tensorflow/tests/embedding_pipelining.mlir
%cst_12 = "tf.Const"() {_replication_info = "repl_info", _xla_compile_device_type = "TPU", device = "", value = dense<1> : tensor<i64>} : () -> tensor<i64> "tf.AssignAddVariableOp"(%37, %cst_12) {_has_manual_control_dependencies = true, _replication_info = "while/cluster_while_body_451", _xla_compile_device_type = "TPU", device = ""} : (tensor<*x!tf_type.resource<tensor<i64>>>, tensor<i64>) -> ()
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Mon Oct 30 06:52:55 UTC 2023 - 33.1K bytes - Viewed (0) -
tensorflow/compiler/mlir/tensorflow/tests/tpu_device_propagation.mlir
// RUN: tf-opt %s -tf-tpu-device-propagation | FileCheck %s // Tests function passthrough values. // CHECK-LABEL: func @testArgToRet // CHECK-SAME: ({{%.+}}: tensor<i64> {tf.device = "/job:localhost/replica:0/task:0/device:TPU:0"}) // CHECK-SAME: -> (tensor<i64> {tf.device = "/job:localhost/replica:0/task:0/device:TPU:0"}) func.func @testArgToRet(%arg0: tensor<i64> {tf.device = "/job:localhost/replica:0/task:0/device:TPU:0"}) -> tensor<i64> {
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Mon Mar 28 12:06:33 UTC 2022 - 19K bytes - Viewed (0) -
tensorflow/compiler/mlir/quantization/tensorflow/quantize_preprocess.cc
// of dropping the attribute. This need not be correct for models not trained // on TPU. // Extracts the StableHLO module from tf.XlaCallModuleOp if the StableHLO // module is serialized in it. pm.addPass(mlir::odml::CreateLegalizeTFXlaCallModuleToStablehloPass()); // Preprocesses TPU-targeting StableHLO module for support in TF Quantizer. pm.addPass(mlir::quant::CreateConvertTpuModelToCpuPass());
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Wed Apr 24 12:49:45 UTC 2024 - 9.8K bytes - Viewed (0) -
tensorflow/compiler/mlir/tf2xla/internal/passes/tpu_cluster_formation.cc
// device. Device attribute is used to infer the appropriate sharding // within TPUs for this op. // TODO(b/183598857): Use explicit sharding ops from the front-end. // For example, dequeue ops generated by // tensorflow/python/tpu/tpu_feed.py if (!tensorflow::IsTPUReplicatedCore(attr.getValue())) { inner_op->removeAttr(kDeviceAttr); }
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu May 02 22:03:30 UTC 2024 - 39.3K bytes - Viewed (0) -
tensorflow/compiler/mlir/tensorflow/transforms/tpu_resource_partitioning.cc
return str_attr && (str_attr.getValue().find("COMPOSITE") != llvm::StringRef::npos); } } // namespace // Rewrites unpartitioned resource reads and writes to partitioned resource // reads and writes. The TPU computation from the frontend is generated in such // a way that resource operations operate on the unpartitioned resource handle // (from a `tf.TPUReplicatedInput`). This results in resource reads and writes
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Wed Sep 06 19:12:29 UTC 2023 - 11.8K bytes - Viewed (0) -
tensorflow/compiler/mlir/tensorflow/transforms/replicate_to_island.cc
// and logical core (`logical_core`). // `replica_id` is the index of the ancestor ReplicateOp in [0, num_replicas). // `logical_core` is the index of the TPU core in [0, num_cores_per_replica). // `device_ordinal` is the index of the TPU core relative to its host. LogicalResult GetDeviceOrdinal(const std::optional<DictionaryAttr>& devices, const unsigned replica_id,
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Mon Jul 24 21:01:40 UTC 2023 - 16.9K bytes - Viewed (0) -
tensorflow/compiler/mlir/tensorflow/transforms/prepare_tpu_computation_for_tf_export.cc
func.getLoc(), op.getOperandTypes(), /*dynamic_key=*/dynamic_key, op.getSendKeyAttr(), /*device_ordinal=*/rewriter.getI64IntegerAttr(0), rewriter.getStringAttr("TPU")); for (auto result : llvm::zip(cloned_func.getArguments(), recv_at_host->getResults())) { std::get<0>(result).replaceAllUsesWith(std::get<1>(result)); }
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 25 16:01:03 UTC 2024 - 11.8K bytes - Viewed (0) -
tensorflow/compiler/mlir/tfrt/tests/mlrt/tpu_conversions.mlir
// RUN: tf-tfrt-opt --split-input-file -pass-pipeline='builtin.module(pre-parallel-tf-to-mlrt{use-tpu-host-allocator-for-inputs=true},tf-mlrt-parallelization{tfrt-cost-threshold=4},tf-to-mlrt)' %s | FileCheck %s --dump-input=fail --dump-input-filter=all func.func @callee(%arg0: tensor<i32>, %arg1: tensor<i32>) -> (tensor<i32>) { func.return %arg0: tensor<i32> } // CHECK-LABEL: func @batch_function func.func @batch_function(%arg0: tensor<i32>) -> (tensor<i32>) {
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Wed Oct 04 21:25:31 UTC 2023 - 11K bytes - Viewed (0) -
tensorflow/compiler/mlir/tf2xla/internal/passes/xla_broadcast.cc
auto status_or_device_coodinates = tensorflow::GetDeviceCoordinates(device_assignment_attr); if (!status_or_device_coodinates.ok()) return cluster.emitError() << "error in fetching tpu device coordinates: " << status_or_device_coodinates.status().message(); status_or_tpu_device_assignment = tensorflow::GetTPUCompilationAndExecutionDevices(
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Jun 13 18:52:07 UTC 2024 - 13.9K bytes - Viewed (0) -
tensorflow/compiler/mlir/tf2xla/internal/passes/mark_ops_for_outside_compilation.cc
RewritePatternSet* patterns) { for (auto op : context->getRegisteredOperations()) op.getCanonicalizationPatterns(*patterns, context); } // Adds the list of ops that are supported on TPU through constant folding which // may depend on the inputs shapes not known at this point. Such ops may not // have any legalization or canonicalization patterns but shouldn't be marked // for outside compilation. //
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 25 16:01:03 UTC 2024 - 21.4K bytes - Viewed (0)