All pull requests

Pull requests

• Bump pytorch-lightning from 1.6.0 to 2.6.1 in /examples/albef

  CLA Signeddependenciespython
  #549·dependabot[bot] opened May 18, 2026Bot·

  ·

• Bump pytorch-lightning from 1.6.0 to 2.6.1 in /examples/mdetr

  CLA Signeddependenciespython
  #548·dependabot[bot] opened May 18, 2026Bot·

  ·

• Bump pillow from 9.3.0 to 12.2.0 in /examples/albef

  CLA Signeddependenciespython
  #547·dependabot[bot] opened May 5, 2026Bot·

  ·

• Bump pillow from 9.3.0 to 12.2.0 in /examples/flava

  CLA Signeddependenciespython
  #546·dependabot[bot] opened May 5, 2026Bot·

  ·

• Fix FLAVA bugs #530 and #533 (attention masks and CLS tokens)

  CLA Signed
  #545·Vishal-sys-code opened Apr 8, 2026·

  ·
  2

• Bump transformers from 4.30.0 to 5.0.0rc3 in /examples/flava

  CLA Signeddependenciespython
  #544·dependabot[bot] opened Apr 8, 2026Bot·

  ·

• Bump transformers from 4.30.0 to 5.0.0rc3 in /examples/albef

  CLA Signeddependenciespython
  #543·dependabot[bot] opened Apr 8, 2026Bot·

  ·

• Bump transformers from 4.30.0 to 5.0.0rc3 in /examples/mugen

  CLA Signeddependenciespython
  #542·dependabot[bot] opened Apr 8, 2026Bot·

  ·

• Bump transformers from 4.30.0 to 5.0.0rc3 in /examples/mdetr

  CLA Signeddependenciespython
  #541·dependabot[bot] opened Apr 8, 2026Bot·

  ·

• Bump requests from 2.31.0 to 2.33.0 in /examples/flava

  CLA Signeddependenciespython
  #540·dependabot[bot] opened Mar 26, 2026Bot·

  ·

• Use scaled_dot_product_attention_with_weights utility

  CLA Signedfb-exportedmeta-exported
  #534·stashuk-olek opened Feb 8, 2026·

  ·
  1

• Update finetune.py

  CLA Signed
  #529·vishwvir-singh opened Sep 6, 2024·

  ·
  2

• Script for comparison of CoCa zero-shot to open_clip implementation

  CLA Signed
  #507·ebsmothers opened Nov 1, 2023Contributor·

  ·
  2

• Add code for LDM

  CLA Signedfb-exported
  #496·abhinavarora opened Oct 17, 2023Contributor·

  ·
  2

• add qformer model to torchmm/models

  CLA Signedfb-exported
  #486·pikapecan opened Oct 11, 2023·

  ·
  7

• add blip2 loss under torchmultimodal/modules/losses

  CLA Signedfb-exported
  #485·pikapecan opened Oct 11, 2023·

  ·
  10

• add blip2 layer under torchmm/models

  CLA Signedfb-exported
  #484·pikapecan opened Oct 11, 2023·

  ·
  9

• create blip2 folder and add qformer layers under torchmm/models

  CLA Signedfb-exported
  #483·pikapecan opened Oct 11, 2023·

  ·
  12

• add triton based Softmax and unit tests

  CLA Signed
  #475·lessw2020 opened Sep 26, 2023Contributor·

  ·
  1

• add Alibi positional embeddings

  CLA Signed
  #462·lessw2020 opened Sep 6, 2023Contributor·

  ·
  14

• Add Parallel_Attention_Blocks (3 of 3)

  CLA Signed
  #457·lessw2020 opened Aug 27, 2023Contributor·

  ·
  18

• Add Rotary Positional Embeddings (RoPE) - part 2 of parallel attention blocks

  CLA Signed
  #450·lessw2020 opened Aug 17, 2023Contributor·

  ·
  16

• Batch matmul fast path in MHAWithCache

  CLA Signedfb-exported
  #449·rohan-varma opened Aug 17, 2023Contributor·

  ·
  9

• Add RMSNorm (part of parallel_attn_blocks)

  CLA Signed
  #448·lessw2020 opened Aug 16, 2023Contributor·

  ·
  16

• Copy transformer impl to oss folder

  CLA Signedfb-exported
  #443·ankitade opened Aug 8, 2023Contributor·

  ·
  3