Commits

Breadcrumbs

History for

onmaster

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 23, 2024

• pnnx fix macos and windows build, add quick test ci (#5838)

  nihui

  authored
  0b71133

Commits on Dec 19, 2024

• pnnx fix onnx clip conversion, add onnx clamp test, always reserve onnx split outputs, convert onnx mod (#5834)

  nihui

  authored
  78aca5d

• pnnx rename _export() to export() (#5831)

  whyb

  authored
  24908b7

• pnnx match onnx zeros ones (#5832)

  nihui

  authored
  57fac4a

Commits on Dec 18, 2024

• pnnx build silent warnings from torch header (#5828)

  nihui

  authored
  ad6d84e

• pnnx auto inputshape from traced inputs (#5825)

  nihui

  authored
  4b0d2de

• pnnx ncnn convert select to crop and squeeze (#5826)

  nihui

  authored
  a12baae

Commits on Nov 27, 2024

• pnnx pass level2 priority (#5791)

  Show description for 1172b04
  nihui

  authored
  1172b04

Commits on Nov 22, 2024

• spectrogram and inverse spectrogram (#5779)

  Show description for 0734b65
  nihui

  authored
  0734b65

• disable pnnx debug model (#5784)

  LJoson

  authored
  c043612

Commits on Nov 7, 2024

• pnnx write implicit int conversion in python script (#5767)

  nihui

  authored
  e71fdf8

Commits on Oct 22, 2024

• pnnx do not fold tensor with dynamic shape, use fp32 module by default (#5755)

  nihui

  authored
  6077adc

Commits on Oct 21, 2024

• pnnx torch 2.5 (#5748)

  nihui

  authored
  c1f9e95

Commits on Oct 15, 2024

• pnnx drop onnx weight-like graph input (#5736)

  nihui

  authored
  f856011

• multiheadattention int8 quantization (#5733)

  Show description for 66b54cb
  nihui

  authored
  66b54cb

Commits on Oct 12, 2024

• gemm int8 quantization (#5706)

  Show description for 1c7af00
  nihui

  authored
  1c7af00

Commits on Sep 4, 2024

• pnnx fuse t5-layernorm as rmsnorm (#5675)

  nihui

  authored
  80c78a0

• update modelwriter for rmsnorm (#5676)

  nihui

  authored
  21e54d8

Commits on Sep 2, 2024

• embed int8 quantization and add embed test (#5667)

  nihui

  authored
  5df5413

Commits on Aug 30, 2024

• pnnx fuse mobilevit style selfattention, onnx2pnnx handle more general gemm (#5659)

  nihui

  authored
  5e2d56d

Commits on Aug 17, 2024

• onnx2pnnx do not fold single constant for gemm weight (#5634)

  nihui

  authored
  4de5369

Commits on Aug 16, 2024

• pnnx functionize do not create shadow op for identity consumers (#5632)

  nihui

  authored
  789d868

Commits on Aug 15, 2024

• RMSNorm (#5630)

  nihui

  authored
  fdf0df3

Commits on Aug 14, 2024

• pnnx drop torch.max torch.min indice node if not used (#5629)

  nihui

  authored
  abad90c

• pnnx convert nn.RMSNorm F.rms_norm (#5628)

  nihui

  authored
  eb6e084

• pnnx convert onnx resize with roi, torch.max torch.min with dim returns tuple (#5627)

  Show description for c46278d
  nihui

  authored
  c46278d

Commits on Aug 12, 2024

• pnnx2ncnn convert torch.roll with one or two shifts (#5623)

  nihui

  authored
  ecfd88a

• pnnx2ncnn handle F.maxpool without dilation param (#5622)

  nihui

  authored
  f3cd4c2

Commits on Aug 10, 2024

• pnnx ci for torch 2.4 (#5618)

  Show description for b9debee
  nihui

  authored
  b9debee

Commits on Aug 8, 2024

• pnnx handles sdpa batch index (#5617)

  nihui

  authored
  60823a8

Commits on Jul 23, 2024

• Add warning and recommend to use pnnx (#5588)

  lll143653

  authored
  d355b6d

Commits on Jul 19, 2024

• pnnx logaddexp (#5598)

  nihui

  authored
  3ee5c18

Commits on Jul 18, 2024

• Update modelwriter.h for mha scale param

  nihui

  authored
  e820158

• pnnx fuse onnx sdpa pattern and ncnn qdim mha fusion (#5589)

  nihui

  authored
  f825d3a

Commits on Jul 15, 2024

• pnnx convert onnx expand/permute/repeat/reshape/select/slice/cat/ceil/chunk/flatten/floor/maximum/minimum/split/squeeze/stack/transpose/unbind/unsqueeze (#5583)

  nihui

  authored
  569617f

Pagination

PreviousNext