Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 9, 2024

• [bug fix] add parents=True (pytorch#2136)

  Show description for 384726f
  

  felipemello1

  and

  Felipe Mello

  authored
  384726f

  Copy full SHA for 384726f

• guard ckpt imports (pytorch#2133)

  Show description for 0f6dc14
  

  felipemello1

  and

  Felipe Mello

  authored
  0f6dc14

  Copy full SHA for 0f6dc14

Commits on Dec 7, 2024

• fix qat_lora_test (pytorch#2131)

  Show description for 06a8379
  

  felipemello1

  and

  Felipe Mello

  authored
  06a8379

  Copy full SHA for 06a8379

Commits on Dec 6, 2024

• update configs (pytorch#2128)

  Show description for 26b2200
  

  felipemello1

  and

  Felipe Mello

  authored
  26b2200

  Copy full SHA for 26b2200

• pass correct arg (pytorch#2127)

  Show description for fef2c80
  

  felipemello1

  and

  Felipe Mello

  authored
  fef2c80

  Copy full SHA for fef2c80

• Update checkpointing directory (pytorch#2074)

  Show description for 424ffc3
  

  3 peopleauthored
  424ffc3

  Copy full SHA for 424ffc3

• Support Early Exit Loss and/or Layer Dropout (pytorch#1076)

  Show description for f8563dd
  

  mostafaelhoushi

  and

  ebsmothers

  authored
  f8563dd

  Copy full SHA for f8563dd

• Reduce logging output for distributed KD (pytorch#2120)

  joecummings

  authored
  f799211

  Copy full SHA for f799211

• update configs (pytorch#2107)

  Show description for 2b1ee6d
  

  felipemello1

  and

  Felipe Mello

  authored
  2b1ee6d

  Copy full SHA for 2b1ee6d

• Llama 3.3 readme updates (pytorch#2125)

  ebsmothers

  authored
  94561a4

  Copy full SHA for 94561a4

• Llama 3.3 70B (pytorch#2124)

  pbontrager

  authored
  bfaa533

  Copy full SHA for bfaa533

Commits on Dec 4, 2024

• Set gloo process group for FSDP with CPU offload (pytorch#2108)

  ebsmothers

  authored
  5eb04cd

  Copy full SHA for 5eb04cd

• Fix bug in loading multimodal datasets and update tests accordingly (pytorch#2110)

  joecummings

  authored
  9b41f49

  Copy full SHA for 9b41f49

Commits on Dec 3, 2024

• Vector Quantized Embeddings (pytorch#2040)

  RdoubleA

  authored
  e9b9ea5

  Copy full SHA for e9b9ea5

• Adding bf16 training for XPU (pytorch#1953)

  Show description for efa91bf
  

  songhappy

  and

  SalmanMohammadi

  authored
  efa91bf

  Copy full SHA for efa91bf

Commits on Nov 28, 2024

• DPO Recipe Doc (pytorch#2091)

  SalmanMohammadi

  authored
  32e265d

  Copy full SHA for 32e265d

• Deprecate SimpoLoss (pytorch#2063)

  SalmanMohammadi

  authored
  0fdc4f0

  Copy full SHA for 0fdc4f0

• DPO Activation Offloading (pytorch#2087)

  SalmanMohammadi

  authored
  5e5a349

  Copy full SHA for 5e5a349

Commits on Nov 27, 2024

• Update build_linux_wheels.yaml - Pass test-infra input params (pytorch#2086)

  atalman

  authored
  67130d9

  Copy full SHA for 67130d9

• Fix nightly tests for qat_lora_fintune_distributed (pytorch#2085)

  andrewor14

  authored
  ecf8d22

  Copy full SHA for ecf8d22

• Error message on packed=True for stack exchange dataset (pytorch#2079)

  joecummings

  authored
  160fd96

  Copy full SHA for 160fd96

• Fail early with packed=True on MM datasets. (pytorch#2080)

  SalmanMohammadi

  authored
  d923234

  Copy full SHA for d923234

Commits on Nov 26, 2024

• [ez] remove stale pytorch version check (pytorch#2075)

  ebsmothers

  authored
  b5d2e63

  Copy full SHA for b5d2e63

• Add ability to shard custom layers for DPO and LoRA distributed (pytorch#2072)

  joecummings

  authored
  b1aecb1

  Copy full SHA for b1aecb1

• Add support for QAT + LoRA (pytorch#1931)

  andrewor14

  authored
  437a8ff

  Copy full SHA for 437a8ff

• MM Docs nits (pytorch#2067)

  SalmanMohammadi

  authored
  d7f8eb0

  Copy full SHA for d7f8eb0

Commits on Nov 25, 2024

• [export] Add exportable position embedding (pytorch#2068)

  larryliu0820

  authored
  5bc03a9

  Copy full SHA for 5bc03a9

• Switch to PyTorch's built-in RMSNorm (pytorch#2054)

  calvinpelletier

  authored
  1450d61

  Copy full SHA for 1450d61

Commits on Nov 23, 2024

• [export] Add exportable attention and kv cache (pytorch#2049)

  larryliu0820

  authored
  f3d8d3c

  Copy full SHA for f3d8d3c

Commits on Nov 22, 2024

• Fix issue in loading mixed precision vocab pruned models during torchtune generation for evaluation (pytorch#2043)

  ifed-ucsd

  authored
  009adaa

  Copy full SHA for 009adaa

• Use hf transfer as default (pytorch#2046)

  Show description for a9aadf5
  

  felipemello1

  and

  Felipe Mello

  authored
  a9aadf5

  Copy full SHA for a9aadf5

Commits on Nov 21, 2024

• Remove deprecated TiedEmbeddingTransformerDecoder (pytorch#2047)

  EmilyIsCoding

  authored
  2d258f4

  Copy full SHA for 2d258f4

• remove default to ignore safetensors (pytorch#2042)

  Show description for fff60d4
  

  felipemello1

  and

  Felipe Mello

  authored
  fff60d4

  Copy full SHA for fff60d4

• feat(cli): allow users to download models from Kaggle (pytorch#2002)

  KeijiBranshi

  authored
  e9fd56a

  Copy full SHA for e9fd56a

Commits on Nov 20, 2024

• CLIP Text Encoder (pytorch#1969)

  calvinpelletier

  authored
  89f935f

  Copy full SHA for 89f935f

Pagination

Previous

Next