All pull requests

Pull requests

• Run pre-commit once in CI and replace the flake8 hook with ruff

  please test
  #1599·jmduarte opened Oct 30, 2026Member·18 of 36 tasks·

  ·

• Apply QKeras weight quantizers to SimpleRNN, LSTM and GRU

  please test
  #1598·dgaytanv opened Oct 28, 2026Contributor·17 tasks done·

  ·
  1

  2

• Add Conv1DTranspose and Conv2DTranspose to the Vivado and Vitis backends

  please test
  #1597·vloncar opened Oct 27, 2026Contributor·19 of 37 tasks·

  ·
  1

  3

• Add reloadable on-chip weights to the Altera backend

  #1595·jmitrevs opened Oct 23, 2026Contributor·13 of 36 tasks·

  ·

• Altera tracing support

  #1594·jmitrevs opened Oct 23, 2026Contributor·11 of 36 tasks·

  ·

• Altera depthwise

  #1593·laurilaatu opened Oct 21, 2026Contributor·1 of 35 tasks·

  ·

• Support Keras 3 and current dependencies in the optimization API

  please test
  #1588·vloncar opened Oct 17, 2026Contributor·17 of 37 tasks·

  ·
  16

• Add indexed_geometry contrib layers for non-standard pixel geometries

  featureplease test
  #1563·mmolin11 opened Sep 25, 2026·19 of 24 tasks·

  ·
  3

• XLS: support custom accumulator precision

  #1554·vasdommes opened Sep 18, 2026Contributor·18 of 36 tasks·

  ·
  4

• Streamed Multi-Head Attention for Altera Backend

  #1551·bugracyln opened Sep 16, 2026·6 of 12 tasks·

  ·
  1

• Add InstanceNormalization layer support (Keras, Vivado/Vitis)

  #1545·linsen458-spec opened Sep 10, 2026·5 of 8 tasks·

  ·
  1

  4

• Add runtime-selectable weights for BramFactor parameters

  #1544·dimdano opened Sep 8, 2026Contributor·8 tasks done·

  ·
  1

• Softmax cleanup in general, update Altera backend to have per-layer lookups

  #1542·jmitrevs opened Sep 8, 2026Contributor·8 tasks done·

  ·
  8

• Add Conv1D dilation support for io_parallel

  please test
  #1541·AaravGaurdev opened Sep 6, 2026·8 tasks done·

  ·
  5

• patch einsum, leakyrelu name

  bugfixplease test
  #1538·brightneedle opened Aug 31, 2026·1 task done·

  ·
  4

• Retire Quartus backend

  #1537·jmitrevs opened Aug 28, 2026Contributor·8 tasks done·

  ·
  13

• Use HGQ's trained softmax lookup tables in the generated HLS

  bugfixplease test
  #1531·JanFSchulte opened Aug 24, 2026Contributor·18 of 37 tasks·

  ·
  1

  1

• CI: route Bambu tests to a dedicated image

  #1511·nghielme opened Jul 23, 2026Contributor·6 of 7 tasks·

  ·
  2

• Add NanoXplore NG-ULTRA accelerator backend

  keep-open
  #1510·nghielme opened Jul 23, 2026Contributor·7 tasks done·

  ·
  1

• Add Bambu (PandA) HLS backend

  keep-open
  #1509·nghielme opened Jul 23, 2026Contributor·7 tasks done·

  ·
  3

• fix(pytorch): place unsqueeze axis at the correct position

  bugfixplease test
  #1496·areporeporepo opened Jun 29, 2026·

  ·
  5

• HGQ QMHA and QLinformer support for oneAPI

  please testStale
  #1432·laurilaatu opened Feb 9, 2026Contributor·7 tasks done·

  ·
  21

• Sotfmax bw inference fix

  bugfixplease test
  #1428·calad0i opened Jan 30, 2026Contributor·2 tasks done·

  ·

  · v1.3.0
  2

• Add support for parsing simple brevitas layers as part of pytorch models

  keep-openplease testreferencewont-merge
  #1019·JanFSchulte opened Jun 7, 2024Contributor·7 tasks done·

  ·
  14