Commits

Branch selector

 master

User selector

glistening

Datepicker

All time

Commit History

Commits on Jan 18, 2021

• [onert] Introduce depthwise conv2d int8 quantization (Samsung#5723)

  Show description for b1f52ab
  glistening

  authored
  b1f52ab

  Copy full SHA for b1f52ab

Commits on Jan 14, 2021

• [onert] Move GetQuantizedMultipliersAndShifts out of Conv (Samsung#5678)

  Show description for 9cc517a
  glistening

  authored
  9cc517a

  Copy full SHA for 9cc517a

• [onert] Make depthwise_conv2d tests int8-quantization ready (Samsung#5676)

  Show description for b806013
  glistening

  authored
  b806013

  Copy full SHA for b806013

Commits on Jan 12, 2021

• [onert] Support Conv2D int8 channel-wise quantization (Samsung#5619)

  Show description for 207f8ee
  glistening

  authored
  207f8ee

  Copy full SHA for 207f8ee

Commits on Jan 11, 2021

• [onert] CircleGen allows zeropoints and scales in array (Samsung#5614)

  Show description for eee4c4f
  glistening

  authored
  eee4c4f

  Copy full SHA for eee4c4f

Commits on Jan 7, 2021

• [onert] Introduce zero_points and scales in ITensor (Samsung#5591)

  Show description for 709e118
  glistening

  authored
  709e118

  Copy full SHA for 709e118

Commits on Jan 5, 2021

• [onert] throws error when zero_point is out of int32 range (Samsung#5547)

  Show description for de8d32d
  glistening

  authored
  de8d32d

  Copy full SHA for de8d32d

Commits on Jan 4, 2021

• [onert] base_loader loads channel-wise quantization params (Samsung#5544)

  Show description for b7bbbe6
  glistening

  authored
  b7bbbe6

  Copy full SHA for b7bbbe6

• [tflitefile_tool] select_operator will handle quantized_dimension (Samsung#5543)

  Show description for cf681eb
  glistening

  authored
  cf681eb

  Copy full SHA for cf681eb

• [onert] Rename offset to zero_point of TypeInfo and makes CWQ ready (Samsung#5538)

  Show description for 41f06e2
  glistening

  authored
  41f06e2

  Copy full SHA for 41f06e2

Commits on Dec 31, 2020

• [onert] Make base_loader to Sparsity and Quantization in similar way (Samsung#5536)

  Show description for 6784e79
  glistening

  authored
  6784e79

  Copy full SHA for 6784e79

Commits on Dec 30, 2020

• [onert] Support Sub int8 quantization (Samsung#5512)

  Show description for 8424dec
  glistening

  authored
  8424dec

  Copy full SHA for 8424dec

• [onert] Support Add int8 quantization (Samsung#5509)

  Show description for 2c71236
  glistening

  authored
  2c71236

  Copy full SHA for 2c71236

• [onert] Support Mul int8 quantization (Samsung#5495)

  Show description for 3ed35ba
  glistening

  authored
  3ed35ba

  Copy full SHA for 3ed35ba

Commits on Dec 29, 2020

• [onert] Make BinaryArithmetic functions int8 ready (Samsung#5488)

  Show description for 82f810b
  glistening

  authored
  82f810b

  Copy full SHA for 82f810b

Commits on Dec 28, 2020

• [onert] Remove Quant8 suffix from BinaryArithmeticOps (Samsung#5485)

  Show description for 281425b
  glistening

  authored
  281425b

  Copy full SHA for 281425b

• [onert] Rename Tensor::data_offset to data_zero_point (Samsung#5486)

  Show description for 9f22837
  glistening

  authored
  9f22837

  Copy full SHA for 9f22837

Commits on Dec 24, 2020

• [onert] Change CalculateActivationRangeUint8 to cover Int8 type (Samsung#5476)

  Show description for 47e28bd
  glistening

  authored
  47e28bd

  Copy full SHA for 47e28bd

Commits on Dec 22, 2020

• [onert] Support quantize operator (int8 to uint8) (Samsung#5470)

  Show description for 490aa3c
  glistening

  authored
  490aa3c

  Copy full SHA for 490aa3c

• [nnpkg_run] Allows NNFW_TYPE_TENSOR_QUANT8_ASYMM_SIGNED input type (Samsung#5460)

  Show description for 6cdc195
  glistening

  authored
  6cdc195

  Copy full SHA for 6cdc195

• [gen_golden] Support int8 quantization random input (Samsung#5453)

  Show description for c9725d7
  glistening

  authored
  c9725d7

  Copy full SHA for c9725d7

• [onert] Support quantize operator (uint8 to int8) (Samsung#5446)

  Show description for e3ebfd5
  glistening

  authored
  e3ebfd5

  Copy full SHA for e3ebfd5

Commits on Dec 21, 2020

• [nnpkg_run, gen_golden] Support int8 input and output (Samsung#5444)

  Show description for 2542004
  glistening

  authored
  2542004

  Copy full SHA for 2542004

• [onert] Fix typo in ElementwiseUnaryLayer::configure error message (Samsung#5445)

  Show description for c8874b4
  glistening

  authored
  c8874b4

  Copy full SHA for c8874b4

Commits on Dec 17, 2020

• [nnpkgrun] add bufsize_for int8 qasym type (Samsung#5398)

  Show description for 8fb5212
  glistening

  authored
  8fb5212

  Copy full SHA for 8fb5212

Commits on Dec 16, 2020

• [onert] Tidy onert logging (Samsung#5354)

  Show description for dde26c3
  glistening

  authored
  dde26c3

  Copy full SHA for dde26c3

Commits on Dec 15, 2020

• [nnpkg_run] Allows int8 asymmetric quantization output type (Samsung#5358)

  Show description for fa48ff2
  glistening

  authored
  fa48ff2

  Copy full SHA for fa48ff2

Commits on Dec 14, 2020

• [onert] add indentation for each subgraph in StaticShapeInferer (Samsung#5334)

  Show description for 41457ad
  glistening

  authored
  41457ad

  Copy full SHA for 41457ad

Commits on Dec 10, 2020

• [onert] Use fixed-width prefix for logging (Samsung#5318)

  Show description for 4f2d68e
  glistening

  authored
  4f2d68e

  Copy full SHA for 4f2d68e

• [onert] add prefix for multiple line log from Lower (Samsung#5335)

  Show description for 4b5a9db
  glistening

  authored
  4b5a9db

  Copy full SHA for 4b5a9db

Commits on Nov 27, 2020

• [onert] Introduce MEAN cpu kernel specialized for axis=1,2 (Samsung#5178)

  Show description for 1ad8183
  glistening

  authored
  1ad8183

  Copy full SHA for 1ad8183

Commits on Nov 23, 2020

• [onert] Support LeakyReLU in cpu backend (Samsung#5100)

  Show description for f503369
  glistening

  authored
  f503369

  Copy full SHA for f503369

• [onert] Support SQRT in cpu backend (Samsung#5099)

  Show description for c72c486
  glistening

  authored
  c72c486

  Copy full SHA for c72c486

Commits on Nov 20, 2020

• [onert] Support SQUARE in cpu backend (Samsung#5094)

  Show description for 08607e5
  glistening

  authored
  08607e5

  Copy full SHA for 08607e5

• [tflitefile_tool] model_parser will use current environment's python (Samsung#5091)

  Show description for c46ddc0
  glistening

  authored
  c46ddc0

  Copy full SHA for c46ddc0

Pagination

Previous

Next