- Sort Score
- Result 10 results
- Languages All
Results 1 - 10 of 49 for Quantile (0.19 sec)
-
istioctl/pkg/metrics/metrics.go
} return sm, nil } func getLatency(promAPI promv1.API, workloadName, workloadNamespace string, duration time.Duration, quantile float64) (time.Duration, error) { latencyQuery := fmt.Sprintf(`histogram_quantile(%f, sum(rate(%s_bucket{%s=~"%s.*", %s=~"%s.*",reporter="destination"}[%s])) by (le))`, quantile, reqDur, destWorkloadLabel, workloadName, destWorkloadNamespaceLabel, workloadNamespace, duration)
Registered: Fri Jun 14 15:00:06 UTC 2024 - Last Modified: Sat Apr 13 05:23:38 UTC 2024 - 8.4K bytes - Viewed (0) -
tensorflow/compiler/mlir/quantization/stablehlo/tests/passes/quantize/quantize.mlir
// RUN: stablehlo-quant-opt %s -split-input-file -stablehlo-quantize -verify-each=false | FileCheck %s // Tests for PopulateFusedGemmStylePatterns are handled in // quantize_composite_functions for module-level evaluation of functions. module attributes {tf_saved_model.semantics} { // CHECK: quantize_simple_xla_call_module(%[[ARG_0:.+]]: tensor<1x4xf32>) func.func private @quantize_simple_xla_call_module(%arg0: tensor<1x4xf32>) -> tensor<1x3xf32> {
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 18 01:38:40 UTC 2024 - 6.3K bytes - Viewed (0) -
tensorflow/compiler/mlir/quantization/tensorflow/tests/quantize.mlir
// RUN: tf-quant-opt %s -split-input-file -quant-lift-quantizable-spots-as-functions -quant-quantize -verify-each=false | FileCheck %s func.func private @conv(%input: tensor<1x3x4x3xf32> {tf._user_specified_name = "input_tensor"}) -> tensor<*xf32> attributes {tf._construction_context = "kEagerRuntime", tf._input_shapes = [#tf_type.shape<1x3x4x3>]} { %weight = arith.constant dense_resource<__elided__> : tensor<2x3x3x2xf32>
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Wed May 08 19:32:28 UTC 2024 - 6.4K bytes - Viewed (0) -
tensorflow/compiler/mlir/quantization/stablehlo/passes/quantize.cc
patterns.add<StableHloQuantization, StableHloQuantizationReverse>(&ctx); PopulateCommonQuantizationPatterns(ctx, patterns, enable_per_channel_quantized_weight_); // Quantize all quantizable ops, including ops that are not compute-heavy. PopulateAllQuantizablePatterns(ctx, patterns); if (failed(applyPatternsAndFoldGreedily(module_op, std::move(patterns)))) {
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Mon Apr 22 07:08:19 UTC 2024 - 5K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/tests/quantize-dynamic-range-float16.mlir
// RUN: tf-opt %s -tfl-prepare-quantize-dynamic-range="enable-float16-quantization" -tfl-quantize="enable-dynamic-range-quantization=true" | FileCheck --check-prefix=CHECK %s // CHECK-LABEL: QuantizeUnidirectionalLstm func.func @QuantizeUnidirectionalLstm(%arg0: tensor<1x2x3xf32>) -> (tensor<1x2x3xf32>) { %1 = "tfl.pseudo_const"() {value = dense<[[0.1]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu May 02 09:41:17 UTC 2024 - 4.6K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/utils/fake_quant_utils.h
// dequantize ops, and insert them between the tf.FakeQuantWithMinMaxVarsOp // and its users. Value value = tf_op.getOutputs(); auto quantize = rewriter.create<TFL::QuantizeOp>( tf_op.getLoc(), qtype.getValue(), value, qtype); auto dequantize = rewriter.create<TFL::DequantizeOp>( tf_op.getLoc(), res_type, quantize.getOutput());
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 25 16:01:03 UTC 2024 - 6.6K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/tests/default_quant_params.mlir
// RUN: tf-opt %s --tfl-default-quant --tfl-quantize | FileCheck %s // CHECK-LABEL: hardcode_all func.func @hardcode_all(%arg0: tensor<2x2xf32>, %arg1: tensor<2x1xf32>) -> tensor<2x2xf32> { %0 = "tfl.add"(%arg0, %arg1) {fused_activation_function="NONE"}: (tensor<2x2xf32>, tensor<2x1xf32>) -> tensor<2x2xf32> func.return %0 : tensor<2x2xf32> // CHECK: %[[q0:.*]] = "tfl.quantize"(%arg1) <{qtype = tensor<2x1x!quant.uniform<u8:f32, 0.0078431372549019607:128>>}>
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu May 02 09:41:17 UTC 2024 - 8.8K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/quantization/tensorflow/tf_to_quant.cc
return "Legalize TF to quant ops dialect"; } }; // Inserts a "tfl.quantize" and "tfl.dequantize" op pair (QDQs) after the // "tf.FakeQuantWithMinMaxVarsOp" to be constant folded. Since the constant // folding logic will use a "arith.constant" op to replace the // "tf.FakeQuantWithMinMaxVarsOp", the "tfl.quantize" op is used to preserve // the quantization parameters as a TypeAttr and "tfl.dequantize" op used to
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Thu Apr 25 16:01:03 UTC 2024 - 8.1K bytes - Viewed (0) -
tensorflow/compiler/mlir/lite/transforms/quantize_patterns.td
include "tensorflow/compiler/mlir/lite/ir/tfl_ops.td" // Quantize attribute $0 by using quantization parameter from %1. def QuantizeByQuantizedType : NativeCodeCall<"quant::Quantize($0, $1.getValue())">; def F32ElementsAttr : ElementsAttrBase< CPred<"$_self.cast<ElementsAttr>().getShapedType().getElementType().isF32()">, "float constant tensor">; // Squash tfl.dequantize and tfl.quantize pairs.
Registered: Sun Jun 16 05:45:23 UTC 2024 - Last Modified: Tue May 28 23:10:13 UTC 2024 - 2.3K bytes - Viewed (0) -
src/internal/trace/gc_test.go
mmuCurve2 := trace.NewMMUCurve(mu) quantiles := []float64{0, 1 - .999, 1 - .99} for window := time.Microsecond; window < time.Second; window *= 10 { mud1 := mmuCurve.MUD(window, quantiles) mud2 := mmuCurve2.MUD(window, quantiles) for i := range mud1 { if !aeq(mud1[i], mud2[i]) { t.Errorf("for quantiles %v at window %v, want %v, got %v", quantiles, window, mud2, mud1) break } } }
Registered: Wed Jun 12 16:32:35 UTC 2024 - Last Modified: Fri May 17 18:48:18 UTC 2024 - 5.3K bytes - Viewed (0)