Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 17, 2025

• Enforce tensor a dtype == tensor b dtype for where.out in facto

  Show description for e1ea74f
  ethansfng

  authored
  e1ea74f

  Copy full SHA for e1ea74f

• Ref implementations interface fixes

  Show description for ed179c0
  DrJessop

  authored
  ed179c0

  Copy full SHA for ed179c0

• Arm backend: Support channels-last input and output

  Show description for 5348ea9
  AdrianLundell

  authored
  5348ea9

  Copy full SHA for 5348ea9

• Rename conv -> conv2d, conv1d_nchw -> conv1d_ncl, conv1d_nhwc -> conv1d_nlc

  Show description for 97e2299
  ethansfng

  authored
  97e2299

  Copy full SHA for 97e2299

• Android use new llm runner deps (#14361)

  Show description for 02b39bf
  kirklandsign

  authored
  02b39bf

  Copy full SHA for 02b39bf

• NXP backend: Add pre-processing pass to fuse Lienar + Add (#14112)

  Show description for b158753
  MartinPavella

  authored
  b158753

  Copy full SHA for b158753

• [Backend Tester] Add pass rate breakdown by parameterization to markdown summary (#14360)

  Show description for eeecd56
  GregoryComer

  authored
  eeecd56

  Copy full SHA for eeecd56

• limit facto to 4000 bytes than numel

  Show description for 7d5c886
  zonglinpeng

  authored
  7d5c886

  Copy full SHA for 7d5c886

• Back out "Add extra logging in CoreML (#13890)"

  Show description for 75cb986
  metascroy

  authored
  75cb986

  Copy full SHA for 75cb986

• aten mode clone dim order op

  Show description for 8d081ed
  Gasoonjia

  authored
  8d081ed

  Copy full SHA for 8d081ed

• limit facto tensor size to random 4000 numel

  Show description for 6fa2062
  zonglinpeng

  authored
  6fa2062

  Copy full SHA for 6fa2062

• [Windows] Reduce trunk model test count (#14348)

  Show description for 16d8109
  GregoryComer

  authored
  16d8109

  Copy full SHA for 16d8109

• Fix eval_llama_qnn script (#14379)

  Show description for f9264f2
  rohansjoshi

  authored
  f9264f2

  Copy full SHA for f9264f2

• Revert "Quantized Softmax Kernel" (#14364)

  Show description for 56659e4
  lucylq

  authored
  56659e4

  Copy full SHA for 56659e4

• Qualcomm AI Engine Direct - Cat Fix (#14325)

  Show description for facf35d
  winskuo-quic

  authored
  facf35d

  Copy full SHA for facf35d

• [multimodal] Allow float32 image input (#14359)

  Show description for bc18834
  larryliu0820

  authored
  bc18834

  Copy full SHA for bc18834

• Arm backend: Split Arm tutorial into ethosu and vgf (#14299)

  Show description for d25c35a
  Erik-Lundell

  authored
  d25c35a

  Copy full SHA for d25c35a

• Split OSS generic ops into their own buck targets

  Show description for 6e63c47
  mcremon-meta

  authored
  6e63c47

  Copy full SHA for 6e63c47

• limited facto tensor size to be less than 4000 bytes

  Show description for c1b7ec5
  zonglinpeng

  authored
  c1b7ec5

  Copy full SHA for c1b7ec5

• solved serde issue when serializing op using list of faketensor as output

  Show description for ed482bd
  Gasoonjia

  authored
  ed482bd

  Copy full SHA for ed482bd

• Remove LlavaRunner (#14356)

  Show description for 9c12e69
  rohansjoshi

  authored
  9c12e69

  Copy full SHA for 9c12e69

Commits on Sep 16, 2025

• Move ReplaceAtenConvolutionWithCadenceConvolutionPass to common section.

  Show description for 108d29d
  eigen-k

  authored
  108d29d

  Copy full SHA for 108d29d

• Add CI for checking file size (#14274)

  Show description for 720aea5
  cccclai

  authored
  720aea5

  Copy full SHA for 720aea5

• Add tensor deps to llm runner buck (#14354)

  kirklandsign

  authored
  b1a41e7

  Copy full SHA for b1a41e7

• Quantized Softmax Kernel

  Show description for 94f62b7
  skrtskrtfb

  authored
  94f62b7

  Copy full SHA for 94f62b7

• Run Llava with MultimodalRunner (#14250)

  Show description for 0f6385e
  rohansjoshi

  authored
  0f6385e

  Copy full SHA for 0f6385e

• StaticAttentionIOManager: optional callback on logits from prefill

  Show description for 3584da9
  sxu

  authored
  3584da9

  Copy full SHA for 3584da9

• Fix buck target for LLM runner util (#14345)

  Show description for 2f21092
  kirklandsign

  authored
  2f21092

  Copy full SHA for 2f21092

• Update main version to 1.1 post branch cut (#14349)

  abhinaykukkadapu

  authored
  a2d2cb3

  Copy full SHA for a2d2cb3

• [Backend Tester] Increase QNN linux runner size (#14200)

  Show description for d42f132
  GregoryComer

  authored
  d42f132

  Copy full SHA for d42f132

• [Backend Tester] Add ARM TOSA flow (#14190)

  Show description for 41644c2
  GregoryComer

  authored
  41644c2

  Copy full SHA for 41644c2

• Qualcomm AI Engine Direct - Rope Fix (#14328)

  Show description for dd7f4d2
  winskuo-quic

  authored
  dd7f4d2

  Copy full SHA for dd7f4d2

• Qualcomm AI Engine Direct - add seq_mse_candidates setting to SmolLM3 and fixed a bug in the graph drawer (#14295)

  Show description for d61dbb9
  DannyYuyang-quic

  authored
  d61dbb9

  Copy full SHA for d61dbb9

• Remove unused broadcast_util include and dep for op__to_dim_order_copy (#14323)

  Show description for a9e265f
  keyprocedure

  authored
  a9e265f

  Copy full SHA for a9e265f

• Remove unused includes from runner (#14338)

  Show description for b80ac3b
  lucylq

  authored
  b80ac3b

  Copy full SHA for b80ac3b

Pagination

Previous

Next