- Sort Score
- Result 10 results
- Languages All
Results 1 - 9 of 9 for lowerings (0.41 sec)
-
tensorflow/compiler/mlir/lite/transforms/lower_static_tensor_list.cc
bool allow_tensorlist_pass_through_; // This flag will control the behavior of setting the batch size one when the // given batch size is None in order to force to proceed the tensor list op // lowerings. bool default_to_single_batch_; }; // Converts tf.Const containing variant of type TensorList to a tensor of // primitive element types. Each of the individual tensor in the list is
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Tue Jun 11 20:00:43 UTC 2024 - 70.7K bytes - Viewed (0) -
tensorflow/compiler/mlir/g3doc/_includes/tf_passes.md
This op is rewritten to generic ops that perform the scale and shift and can operate on non-quantized types. Currently, TF_DequantizeOp is the only op with a lowering that falls in this category. When more lowerings are added (e.g. QuantizeV2Op), they should be added to this pass. ### `-tf-mark-ops-for-outside-compilation`
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Wed Aug 02 02:26:39 UTC 2023 - 96.4K bytes - Viewed (0) -
src/cmd/compile/internal/ssa/_gen/S390X.rules
(Neg64F ...) => (FNEG ...) (Com64 ...) => (NOT ...) (Com(32|16|8) ...) => (NOTW ...) (NOT x) => (XOR (MOVDconst [-1]) x) (NOTW x) => (XORWconst [-1] x) // Lowering boolean ops (AndB ...) => (ANDW ...) (OrB ...) => (ORW ...) (Not x) => (XORWconst [1] x) // Lowering pointer arithmetic (OffPtr [off] ptr:(SP)) => (MOVDaddr [int32(off)] ptr) (OffPtr [off] ptr) && is32Bit(off) => (ADDconst [int32(off)] ptr)
Registered: Wed Jun 12 16:32:35 UTC 2024 - Last Modified: Thu Oct 12 18:09:26 UTC 2023 - 74.3K bytes - Viewed (0) -
src/cmd/compile/internal/ssa/_gen/AMD64.rules
(Neg64F x) => (PXOR x (MOVSDconst <typ.Float64> [math.Copysign(0, -1)])) // Lowering boolean ops (AndB ...) => (ANDL ...) (OrB ...) => (ORL ...) (Not x) => (XORLconst [1] x) // Lowering pointer arithmetic (OffPtr [off] ptr) && is32Bit(off) => (ADDQconst [int32(off)] ptr) (OffPtr [off] ptr) => (ADDQ (MOVQconst [off]) ptr) // Lowering other arithmetic (Ctz64 x) && buildcfg.GOAMD64 >= 3 => (TZCNTQ x)
Registered: Wed Jun 12 16:32:35 UTC 2024 - Last Modified: Tue Mar 12 19:38:41 UTC 2024 - 93.9K bytes - Viewed (0) -
src/cmd/compile/internal/ssa/_gen/PPC64.rules
(Neg(64|32|16|8) ...) => (NEG ...) (Neg(64|32)F ...) => (FNEG ...) (Com(64|32|16|8) x) => (NOR x x) // Lowering boolean ops (AndB ...) => (AND ...) (OrB ...) => (OR ...) (Not x) => (XORconst [1] x) // Merge logical operations (AND x (NOR y y)) => (ANDN x y) (OR x (NOR y y)) => (ORN x y) // Lowering comparisons (EqB x y) => (ANDconst [1] (EQV x y))
Registered: Wed Jun 12 16:32:35 UTC 2024 - Last Modified: Fri Jun 07 19:02:52 UTC 2024 - 53.2K bytes - Viewed (0) -
tensorflow/compiler/mlir/tensorflow/transforms/lower_tf.cc
float half_range = is_unsigned ? 0 : 128; return GetScalarOfType(builder->getF32Type(), half_range); } // Returns reduction indices to use while lowering tf.BiasAddGrad op to tf.Sum // op. DenseIntElementsAttr GetBiasAddGradReductionIndices(int64_t rank, StringAttr data_format,
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 25 16:01:03 UTC 2024 - 74.9K bytes - Viewed (0) -
pkg/controller/replicaset/replica_set_test.go
t.Fatal(err) } // Get the key for the controller rsKey, err := controller.KeyFunc(rsSpec) if err != nil { t.Errorf("Couldn't get key for object %#v: %v", rsSpec, err) } // Lowering expectations should lead to a sync that creates a replica, however the // fakePodControl error will prevent this, leaving expectations at 0, 0 manager.expectations.CreationObserved(logger, rsKey) rsSpec.Status.Replicas = 1
Registered: Sat Jun 15 01:39:40 UTC 2024 - Last Modified: Sat May 04 18:33:12 UTC 2024 - 69.2K bytes - Viewed (0) -
tensorflow/compiler/mlir/tensorflow/tests/lower_tf.mlir
} // Verify SpaceToBatchND with input tensor of element type f16. This test case is derived from 'fourdim_space_to_batch_nd'. It checks the output // tensor shape and element type in a few lines in the resulting lowering. // CHECK-LABEL: space_to_batch_nd_element_type_f16 func.func @space_to_batch_nd_element_type_f16(%input: tensor<3x5x7x10xf16>, %block_shape: tensor<2xi64>, %paddings: tensor<2x2xi64>) -> tensor<?x?x?x10xf16> {
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Fri Jan 05 18:35:42 UTC 2024 - 92K bytes - Viewed (0) -
src/cmd/compile/internal/ssa/rewrite.go
// of this conversion is: // a' = (ADDCC x y) // a = (Select0 a') // b = (CMPconst [0] a) // c = (OR a z) // // Which makes it trivial to rewrite b using a lowering rule. func convertPPC64OpToOpCC(op *Value) *Value { ccOpMap := map[Op]Op{ OpPPC64ADD: OpPPC64ADDCC, OpPPC64ADDconst: OpPPC64ADDCCconst, OpPPC64AND: OpPPC64ANDCC,
Registered: Wed Jun 12 16:32:35 UTC 2024 - Last Modified: Fri Jun 07 19:02:52 UTC 2024 - 64.2K bytes - Viewed (0)