Commits

Branch selector

 main

User selector

All users

All time

Commit History

Commits on Jul 25, 2024

• remove unnecessary guard code related with pytorch versions 1.4.2 ~ 1.7.0 (#32210)

  Show description for f53a5de
  ji-huazhong

  authored
  f53a5de

  Copy full SHA for f53a5de

• [whisper] fix short-form output type (#32178)

  Show description for 5658e74
  sanchit-gandhi

  authored
  5658e74

  Copy full SHA for 5658e74

Commits on Jul 24, 2024

• fix: Replaced deprecated `unittest method` with the correct one (#32198)

  Show description for 85a1269
  Sai-Suraj-27

  authored
  85a1269

  Copy full SHA for 85a1269

• :rotating_light: No more default chat templates (#31733)

  Show description for edd68f4
  Rocketknight1

  authored
  edd68f4

  Copy full SHA for edd68f4

• Support dequantizing GGUF FP16 format (#31783)

  Show description for 1c122a4
  PenutChen

  authored
  1c122a4

  Copy full SHA for 1c122a4

• Fix float8_e4m3fn in modeling_utils (#32193)

  Show description for af0e4b7
  SunMarc

  authored
  af0e4b7

  Copy full SHA for af0e4b7

• Fix resize embedding with Deepspeed (#32192)

  Show description for 1392a68
  zucchini-nlp

  authored
  1392a68

  Copy full SHA for 1392a68

• let's not warn when someone is running a forward (#32176)

  Show description for 8d2534c
  ArthurZucker

  authored
  8d2534c

  Copy full SHA for 8d2534c

• RoPE: relaxed rope validation (#32182)

  Show description for e0182f3
  gante

  authored
  e0182f3

  Copy full SHA for e0182f3

• Remove conversational pipeline tests (#32099)

  Show description for 165116b
  amyeroberts

  authored
  165116b

  Copy full SHA for 165116b

• Update qwen2.md (#32108)

  Show description for 5f4ee98
  ArtificialZeng

  authored
  5f4ee98

  Copy full SHA for 5f4ee98

• fix: default value reflects the runtime environment variables rather than the ones present at import time. (#32153)

  Show description for 8678879
  junrae6454

  authored
  8678879

  Copy full SHA for 8678879

• adds: extra_repr() to MambaRMSNorm to include hidden size / size of weights in the layer (#32171)

  Show description for 01be5b4
  rohitdwivedula

  authored
  01be5b4

  Copy full SHA for 01be5b4

Commits on Jul 23, 2024

• [docs] change temperature to a positive value (#32077)

  Show description for c85510f
  faaany

  authored
  c85510f

  Copy full SHA for c85510f

• fix: Fixed an if condition that is always evaluating to true (#32160)

  Show description for bc2adb0
  Sai-Suraj-27

  authored
  bc2adb0

  Copy full SHA for bc2adb0

• fix (#32162)

  gante

  authored
  23f6a43

  Copy full SHA for 23f6a43

• Llama 3.1 conversion

  Show description for d5a99df
  

  LysandreJik

  and

  ArthurZucker

  committed
  d5a99df

  Copy full SHA for d5a99df

• Dev version: v4.44.0.dev0

  LysandreJik

  committed
  ff0d708

  Copy full SHA for ff0d708

• Updated `ruff` to the latest version (#31926)

  Show description for d2c687b
  Sai-Suraj-27

  authored
  d2c687b

  Copy full SHA for d2c687b

• Enhancing SFT Training Efficiency Using Packing and FlashAttention2 with Position IDs (#31629)

  Show description for 9cf4f2a
  RhuiDih

  authored
  9cf4f2a

  Copy full SHA for 9cf4f2a

• Added additional kwarg for successful running of optuna hyperparameter search (#31924)

  Show description for 7d92009
  DeF0017

  authored
  7d92009

  Copy full SHA for 7d92009

• feat(cache): StaticCache uses index_copy_ to avoid useless copy (#31857)

  Show description for 6370062
  tengomucho

  authored
  6370062

  Copy full SHA for 6370062

• Fix typing to be compatible with later py versions (#32155)

  amyeroberts

  authored
  a009fbd

  Copy full SHA for a009fbd

• Revert "Incorrect Whisper long-form decoding timestamps " (#32148)

  Show description for 3263b34
  sanchit-gandhi

  authored
  3263b34

  Copy full SHA for 3263b34

• Rename Phi-3 rope scaling type (#31436)

  Show description for 034b477
  garg-amit

  authored
  034b477

  Copy full SHA for 034b477

• Added mamba.py backend (#30139)

  Show description for bab32d6
  

  alxndrTL

  and

  ArthurZucker

  authored
  bab32d6

  Copy full SHA for bab32d6

• Fix video batching to videollava (#32139)

  Show description for 9ced33c
  

  merveenoyan

  and

  Merve Noyan

  authored
  9ced33c

  Copy full SHA for 9ced33c

• Fix flash attention speed issue (#32028)

  Show description for a5b226c
  Cyrilvallez

  authored
  a5b226c

  Copy full SHA for a5b226c

• gguf conversion add_prefix_space=None for llama3 (#31937)

  Show description for a1844a3
  itazap

  authored
  a1844a3

  Copy full SHA for a1844a3

• Llama: RoPE refactor (#32135)

  Show description for 2e11342
  

  3 peopleauthored
  2e11342

  Copy full SHA for 2e11342

• Modify resize_token_embeddings to ensure output type is same as input (#31979)

  Show description for 5a4a76e
  

  3 peopleauthored
  5a4a76e

  Copy full SHA for 5a4a76e

• Disable quick init for TapasPreTrainedModel (#32149)

  Show description for 1535a2c
  daniellok-db

  authored
  1535a2c

  Copy full SHA for 1535a2c

• Add YaRN and Dynamic-YaRN RoPE Scaling Methods (#30910)

  Show description for 34b4321
  

  3 peopleauthored
  34b4321

  Copy full SHA for 34b4321

• Add method to retrieve used chat template (#32032)

  Show description for 7405c1c
  KonradSzafer

  authored
  7405c1c

  Copy full SHA for 7405c1c

• Fix mask creations of `GPTNeoX` and `GPT2` (#31944)

  Show description for 605f324
  vasqu

  authored
  605f324

  Copy full SHA for 605f324

Pagination

Previous

Next