Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jun 6, 2025

• Merge branch 'auphelia-fix/io_tensor_order'

  maltanar

  committed
  7bb3a27

  Copy full SHA for 7bb3a27

• [Test] remove time measurement and return value from test

  maltanar

  committed
  5cfacb8

  Copy full SHA for 5cfacb8

• Merge pull request #186 from auphelia/fix/io_tensor_order

  Show description for 0630cea
  maltanar

  authored
  0630cea

  Copy full SHA for 0630cea

• [Test] add unit test checking IO order with shape setting

  maltanar

  committed
  c1173cc

  Copy full SHA for c1173cc

Commits on May 29, 2025

• [ModelWrapper] Ensure that value info tensor gets added

  auphelia

  committed
  dfdb22d

  Copy full SHA for dfdb22d

• [ModelWrapper] Fix set_tensor_shape method to preserve order of inputs and outputs

  auphelia

  committed
  0230d09

  Copy full SHA for 0230d09

Commits on May 21, 2025

• Merge pull request #184 from fastmachinelearning/feature/floatquant_readme_updates

  Show description for 81696c7
  maltanar

  authored
  81696c7

  Copy full SHA for 81696c7

• Merge pull request #183 from fastmachinelearning/feature/remove_coverage

  Show description for f9610c8
  maltanar

  authored
  f9610c8

  Copy full SHA for f9610c8

• Update README.md

  maltanar

  authored
  5f53e4c

  Copy full SHA for 5f53e4c

• disable coverage in pytest

  maltanar

  authored
  68c0c80

  Copy full SHA for 68c0c80

Commits on May 19, 2025

• Merge pull request #182 from fastmachinelearning/feature/arbprec_float_dtype

  Show description for 30f0a23
  maltanar

  authored
  30f0a23

  Copy full SHA for 30f0a23

Commits on May 9, 2025

• Update pre-commit.yml

  maltanar

  authored
  d4674b1

  Copy full SHA for d4674b1

Commits on May 8, 2025

• Merge branch 'main' into feature/arbprec_float_dtype

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  3498894

  Copy full SHA for 3498894

• Merge pull request #180 from ebby-s/feature/arbprec_float_dtype

  Show description for 19c84eb
  maltanar

  authored
  19c84eb

  Copy full SHA for 19c84eb

Commits on Apr 28, 2025

• Added tests for ArbPrecFloatType with custom exponent bias.

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  1611229

  Copy full SHA for 1611229

• ArbPrecFloatType.allowed() works for denormal values in target minifloat, added tests.

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  d47c98f

  Copy full SHA for d47c98f

• ArbPrecFloatType takes optional exp bias, allowed() uses range check and assumes no denormals.

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  e0b227e

  Copy full SHA for e0b227e

Commits on Apr 22, 2025

• Implemented FloatQuant.infer_node_datatype().

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  fe5c85e

  Copy full SHA for fe5c85e

• Added tests for ArbPrecFloat.

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  8d52071

  Copy full SHA for 8d52071

Commits on Apr 11, 2025

• Merge branch 'feature/float_quant' into feature/arbprec_float_dtype

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  61b2f8c

  Copy full SHA for 61b2f8c

• Added a unit test that tests the QONNX execution of an exported FloatQuant graph.

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  d2fcb60

  Copy full SHA for d2fcb60

• Added data for test_float_exported_graph_exec.

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  39eaafa

  Copy full SHA for 39eaafa

Commits on Apr 10, 2025

• Renamed Quant to IntQuant, preserved backwards compatibility for refs to quant.py and Quant.

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  e686f55

  Copy full SHA for e686f55

• Added dependencies (hypothesis, mock, brevitas) for testing.

  

  Ebby Samson

  authored and

  Ebby Samson

  committed
  657cd8c

  Copy full SHA for 657cd8c

• Merge pull request #159 from nghielme/float_quant

  Show description for 7ca4c87
  maltanar

  authored
  7ca4c87

  Copy full SHA for 7ca4c87

Commits on Mar 28, 2025

• [FloatQuant] update documentation and improve float_quant implementation. Default exponent bias is now computed if not provided, and tests have been added to compare QONNX and Brevitas float quanti…

  Show description for ee45e6f
  nghielme

  committed
  ee45e6f

  Copy full SHA for ee45e6f

Commits on Feb 28, 2025

• Merge pull request #172 from makoeppel/python/remove_support_for_38_39

  Show description for ed33937
  maltanar

  authored
  ed33937

  Copy full SHA for ed33937

Commits on Feb 24, 2025

• [FloatQuant] integrate FloatQuant custom operation and refactor float quantization logic. Now QONNX and Brevitas float quantisers match.

  nghielme

  committed
  2e51b8b

  Copy full SHA for 2e51b8b

• Merge branch 'main' into python/remove_support_for_38_39

  makoeppel

  authored
  39385c6

  Copy full SHA for 39385c6

• start to remove python 3.8 / 3.9

  makoeppel

  committed
  1b0b81b

  Copy full SHA for 1b0b81b

Commits on Feb 1, 2025

• Merge pull request #167 from fastmachinelearning/feature/dtinf_no_scaledint

  Show description for 5b0d9fd
  maltanar

  authored
  5b0d9fd

  Copy full SHA for 5b0d9fd

Commits on Jan 30, 2025

• [Test] enable scaled-int dtypes for pruning test for inf cost

  maltanar

  committed
  5b050a7

  Copy full SHA for 5b050a7

• [InfCost] add allow_scaledint_dtypes=True earlier

  maltanar

  committed
  ce063a9

  Copy full SHA for ce063a9

• [InferDT] disable scaled-int dtype inference by default

  maltanar

  committed
  c34bd9a

  Copy full SHA for c34bd9a

Commits on Jan 24, 2025

• [Test] add dtype inference with & without scaled int option

  maltanar

  committed
  2b56701

  Copy full SHA for 2b56701

Pagination

Previous

Next