Search Options

Results per page
Sort
Preferred Languages
Advance

Results 1 - 2 of 2 for 1x2x2x1024xbf16 (0.26 sec)

  2. github.com/tensorflow/tensorflow

    tensorflow/compiler/mlir/quantization/tensorflow/tests/propagate_quantize_type.mlir 

        %2 = "tf.XlaDotV2"(%arg0, %1) {device = "", dimension_numbers = "\12\01\00\0A\01\03", precision_config = ""} : (tensor<1x2x2x2xbf16>, tensor<2x1024xbf16>) -> tensor<1x2x2x1024xbf16>
    %3 = "tf.Identity"(%2) : (tensor<1x2x2x1024xbf16>) -> tensor<1x2x2x1024xbf16>
    return %3 : tensor<1x2x2x1024xbf16>
  }

  func.func private @composite_dequantize_uniform(%arg0: tensor<*xi8>) -> tensor<*xbf16> {
    Registered: Sun Jun 16 05:45:23 UTC 2024
    - Last Modified: Mon Oct 30 06:52:55 UTC 2023
    - 6.6K bytes
    - Viewed (0)
  3. github.com/tensorflow/tensorflow

    tensorflow/compiler/mlir/quantization/tensorflow/tests/quantize_weights.mlir 

    module {
  func.func @quantize_xladotv2_bf16(%arg0: tensor<1x2x2x2xbf16>) -> (tensor<1x2x2x1024xbf16>) {
    %cst_0 = "tf.Const"() {value = dense<2.000000e+00> : tensor<2x1024xbf16>} : () -> tensor<2x1024xbf16>
    %0 = "tf.XlaDotV2"(%arg0, %cst_0) {device = "", dimension_numbers = "\12\01\00\0A\01\03", precision_config = ""} : (tensor<1x2x2x2xbf16>, tensor<2x1024xbf16>) -> tensor<1x2x2x1024xbf16>
    // Check dequantize performed in bf16.
    Registered: Sun Jun 16 05:45:23 UTC 2024
    - Last Modified: Mon Oct 30 06:52:55 UTC 2023
    - 42K bytes
    - Viewed (0)
Back to top