Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Feb 2, 2025

• #16948: Add fixes to support unpadded inner dim for pf + mm and add support for DRAM in1 inputs for ring matmul. (#17311)

  Show description for 5db3a16
  avoraTT

  authored
  5db3a16

  Copy full SHA for 5db3a16

• 0: Fix Packed unicast size

  Show description for 69592bb
  jbaumanTT

  committed
  69592bb

  Copy full SHA for 69592bb

• LightMetal - Use as_*() instead of static_cast directly for some union type handling in from_flatbuffer()

  Show description for 36df59d
  kmabeeTT

  committed
  36df59d

  Copy full SHA for 36df59d

• LightMetal - Add flatbuffer schema and conversion to/from functions for various types (#17039)

  Show description for 152b88b
  kmabeeTT

  committed
  152b88b

  Copy full SHA for 152b88b

• #12253: Batch Norm support for training mode (#16592)

  Show description for b5e983b
  

  VirdhatchaniKN

  and

  mouliraj-mcw

  authored
  b5e983b

  Copy full SHA for b5e983b

• [UMD] Dram view in soc descriptor (#17306)

  Show description for ed7c874
  broskoTT

  authored
  ed7c874

  Copy full SHA for ed7c874

• #16704: Added memory accounting for Bias in Lin Op (#17056)

  Show description for 79a6913
  vsureshTT

  authored
  79a6913

  Copy full SHA for 79a6913

• Initial SubDevice support for TT-Mesh + tests

  tt-asaigal

  committed
  2e80403

  Copy full SHA for 2e80403

• #17324: port Padding CPP Unit OP Tests to TTNN (#17354)

  Show description for 32b9207
  llongTT

  authored
  32b9207

  Copy full SHA for 32b9207

• #17203: Remove work executor from MeshDevice (#17475)

  Show description for 0a86356
  omilyutin-tt

  authored
  0a86356

  Copy full SHA for 0a86356

Commits on Feb 1, 2025

• #0: Remove TT_ASSERT statement that break tt-train in debug mode (#17472)

  Show description for 1fe0579
  omilyutin-tt

  authored
  1fe0579

  Copy full SHA for 1fe0579

• #17203: Remove 2 worker-queue related methods from IDevice (#17317)

  Show description for d8d3e54
  omilyutin-tt

  authored
  d8d3e54

  Copy full SHA for d8d3e54

• [skip ci] Update metal-api-surface.yaml (#17470)

  blozano-tt

  authored
  b874abf

  Copy full SHA for b874abf

• #16871: Unify conditional typecasting in BinaryNg (#17148)

  Show description for 55dd492
  

  mcw-anasuya

  and

  KalaivaniMCW

  authored
  55dd492

  Copy full SHA for 55dd492

• [skip ci] Compute the surface of tt_metal API (#17469)

  Show description for 2877adf
  blozano-tt

  authored
  2877adf

  Copy full SHA for 2877adf

• Extra cleanup. With emphasis (#17466)

  Show description for f8fe02d
  afuller-TT

  authored
  f8fe02d

  Copy full SHA for f8fe02d

• #17203: Resolve a deadlock resulting in graph tracker tracking allocation before it was completed (#17436)

  Show description for 4078e26
  omilyutin-tt

  authored
  4078e26

  Copy full SHA for 4078e26

• Optimized launch_op compilation (#17448) Tested locally only :(

  Show description for f30bfcb
  dmakoviichuk-tt

  authored
  f30bfcb

  Copy full SHA for f30bfcb

Commits on Jan 31, 2025

• [skip ci] Update vLLM commit for deepseek-distill-llama70b in README.md (#17459)

  skhorasganiTT

  authored
  70206b9

  Copy full SHA for 70206b9

• Rename ttnn::SimpleShape to ttnn::Shape (#17454)

  Show description for 2df6e60
  sminakov-tt

  authored
  2df6e60

  Copy full SHA for 2df6e60

• Update the .rst to match code (#17455)

  Show description for b6d72ad
  afuller-TT

  authored
  b6d72ad

  Copy full SHA for b6d72ad

• Remove ttnn::Shape and LegacyShape from the codebase (#17413)

  Show description for 8063f0a
  sminakov-tt

  authored
  8063f0a

  Copy full SHA for 8063f0a

• #0: Modify DeviceRange semantics to match CoreRange

  Show description for f9b52ff
  tt-asaigal

  committed
  f9b52ff

  Copy full SHA for f9b52ff

• #0: fix bug in createqkv (#17452)

  Show description for c7825a1
  kpaigwar

  authored
  c7825a1

  Copy full SHA for c7825a1

• [skip ci] #0: Add failure signature for runner shutdown (#17439)

  Show description for e106618
  williamlyTT

  authored
  e106618

  Copy full SHA for e106618

• [Llama3.2-11b-vision] Add max_cross_attn_tokens property to vLLM generator class (#17401)

  skhorasganiTT

  authored
  2b29d50

  Copy full SHA for 2b29d50

• [skip ci] Add DeepSeek perf and instructions (#17445)

  Show description for 12efa1f
  yieldthought

  authored
  12efa1f

  Copy full SHA for 12efa1f

• Step 1 of normalizing Boost as a dependency (#17446)

  Show description for ed0d2e1
  afuller-TT

  authored
  ed0d2e1

  Copy full SHA for ed0d2e1

• Add Qwen2.5 vLLM generator (based on LlamaGenerator), fix batch 1 issue with generator's decode forward (#17422)

  skhorasganiTT

  authored
  41d4b36

  Copy full SHA for 41d4b36

• Address clang tidy error (#17437)

  Show description for 47fe29f
  afuller-TT

  authored
  47fe29f

  Copy full SHA for 47fe29f

• #16380: add nD support for generic reductions (#17399)

  Show description for c109516
  bbradelTT

  authored
  c109516

  Copy full SHA for c109516

• #0: Update pgm_dispatch_golden.json

  Show description for 6c87f4c
  jbaumanTT

  committed
  6c87f4c

  Copy full SHA for 6c87f4c

• #0: Change noc_inline_dw_write to use write_at_cmd_buf to avoid overwriting cached register values of write_reg_cmd_buf

  tt-aho

  committed
  d354e26

  Copy full SHA for d354e26

• #15414: Read annotation data to determine job-level failure signature and reason (#17410)

  Show description for e099072
  williamlyTT

  authored
  e099072

  Copy full SHA for e099072

• Xuncai/all gather fixes (#17155)

  Show description for a53f8dd
  

  caixunshiren

  and

  avoraTT

  authored
  a53f8dd

  Copy full SHA for a53f8dd

Pagination

Previous

Next