All pull requests

Pull requests

• Register l1_pruning and fpgm with kind="pruning"

  #37·dip-era opened Nov 4, 2026·

  ·
  1

• Fix pruning crash on CUDA when a layer feeds a Linear

  #36·dip-era opened Nov 4, 2026·

  ·
  1

• Feat/add ptq quantization

  #32·Rusty-Almond opened Oct 29, 2026Contributor·

  ·
  1

• Fix/serialization modules

  #31·vishweshkhandareiitb-creator opened Oct 29, 2026·

  ·
  1

• Add Pipeline class to combine compression methods

  #30·sumedhabhattacharya0711-spec opened Oct 25, 2026·

  ·
  1

• Fix/hrank pruning shapes

  #29·vishweshkhandareiitb-creator opened Oct 24, 2026·

  ·
  1

• Feature/add onnx export

  #28·vishweshkhandareiitb-creator opened Oct 21, 2026·

  ·
  1

• Feat/add options to finetune

  #27·Rusty-Almond opened Oct 21, 2026Contributor·

  ·
  1

• Allow a different pruning ratio for each layer (#5)

  #26·Abishekcps opened Oct 20, 2026·

  ·
  1

• Improve knowledge distillation compression ( ISSUE #1)

  #25·Tarun-zz opened Oct 18, 2026·

  ·
  1

• add memory estimation for weights, activations, and peak memory (#3)

  #24·vishweshkhandareiitb-creator opened Oct 18, 2026·

  ·
  1

• Support grouped pruning for layers that share channels (residual adds)

  #22·bhattyuvraj22 opened Oct 16, 2026Contributor·

  ·
  1

• Fix quantized engine restore failing when default engine is 'none'

  #16·bhattyuvraj22 opened Oct 10, 2026Contributor·

  ·
  1

• fix(quantization): modernize DFQ kernels and guard deprecated torch.ao APIs (#7)

  #15·gaurika-analyst opened Oct 10, 2026·

  ·
  1

  2