Skip to content

GitLab

Switch branch/tag
  1. 23 Mar, 2022 1 commit
  3. 22 Mar, 2022 4 commits
  5. 21 Mar, 2022 9 commits
  7. 18 Mar, 2022 1 commit
  9. 17 Mar, 2022 2 commits
  11. 16 Mar, 2022 1 commit
  13. 15 Mar, 2022 4 commits
  15. 14 Mar, 2022 4 commits
  17. 12 Mar, 2022 1 commit
    • Stas Bekman's avatar
      [Deepspeed] add support for bf16 mode (#14569) · 580dd87c
      Stas Bekman authored 
      

* [WIP] add support for bf16 mode

* prep for bf16

* prep for bf16

* fix; zero2/bf16 is ok

* check bf16 is available

* test fixes

* enable zero3_bf16

* config files

* docs

* split stage_dtype; merge back to non-dtype-specific config file

* fix doc

* cleanup

* cleanup

* bfloat16 => bf16 to match the PR changes

* s/zero_gather_fp16_weights_on_model_save/zero_gather_16bit_weights_on_model_save/; s/save_fp16_model/save_16bit_model/

* test fixes/skipping

* move

* fix

* Update docs/source/main_classes/deepspeed.mdx
Co-authored-by: default avatarSylvain Gugger <35901082+sgugger@users.noreply.github.com>

* backticks

* cleanup

* cleanup

* cleanup

* new version

* add note about grad accum in bf16
Co-authored-by: default avatarSylvain Gugger <35901082+sgugger@users.noreply.github.com>
      580dd87c
  19. 11 Mar, 2022 3 commits
  21. 10 Mar, 2022 3 commits
  23. 09 Mar, 2022 3 commits
    • Sanchit Gandhi's avatar
      Add FlaxBartForCausalLM (#15995) · b256f351
      Sanchit Gandhi authored 
      * add causal lm

* add CausalLM tests

* Add FlaxBartForCausalLM

* Add EncoderDecoder model tests

* change docstring

* make repo-consistency

* suggested changes

* remove jax ops

* correction

* rename pre-trained decoder model
      b256f351
    • lewtun's avatar
      Add ONNX export for ViT (#15658) · 50dd314d
      lewtun authored 
      

* Add ONNX support for ViT

* Refactor to use generic preprocessor

* Add vision dep to tests

* Extend ONNX slow tests to ViT

* Add dummy image generator

* Use model_type to determine modality

* Add deprecation warnings for tokenizer argument

* Add warning when overwriting the preprocessor

* Add optional args to docstrings

* Add minimum PyTorch version to OnnxConfig

* Refactor OnnxConfig class variables from CONSTANT_NAME to snake_case

* Add reasonable value for default atol
Co-authored-by: default avatarSylvain Gugger <35901082+sgugger@users.noreply.github.com>
      50dd314d
    • Patrick von Platen's avatar
      [Doctests] Move doctests to new GPU & Fix bugs (#15969) · c1aaa439
      Patrick von Platen authored 
      

* test

* up

* up

* Empty test commit

* up

* update tests

* up

* fix some vision models

* correct

* correct docs

* Trigger notification

* finalize

* check

* correct quicktour

* Apply suggestions from code review

* improve doctests

* Trigger Build

* next try

* next try

* and again

* Output current clone information

* Output current clone information

* Correct path

* add tf round again

* revert to daily job
Co-authored-by: default avatarLysandre <lysandre.debut@reseau.eseo.fr>
      c1aaa439
  25. 07 Mar, 2022 1 commit
  27. 04 Mar, 2022 3 commits
    • Chan Woo Kim's avatar
      Constrained Beam Search [*With* Disjunctive Decoding] (#15761) · 5c6f57ee
      Chan Woo Kim authored 
      

* added classes to get started with constrained beam search

* in progress, think i can directly force tokens now but not yet with the round robin

* think now i have total control, now need to code the bank selection

* technically works as desired, need to optimize and fix design choices leading to undersirable outputs

* complete PR #1 without disjunctive decoding

* removed incorrect tests

* Delete k.txt

* Delete test.py

* Delete test.sh

* revert changes to test scripts

* genutils

* full implementation with testing, no disjunctive yet

* shifted docs

* passing all tests realistically ran locally

* removing accidentally included print statements

* fixed source of error in initial PR test

* fixing the get_device() vs device trap

* fixed documentation docstrings about constrained_beam_search

* fixed tests having failing for Speech2TextModel's floating point inputs

* fix cuda long tensor

* added examples and testing for them and founx & fixed a bug in beam_search and constrained_beam_search

* deleted accidentally added test halting code with assert False

* code reformat

* Update tests/test_generation_utils.py
Co-authored-by: default avatarPatrick von Platen <patrick.v.platen@gmail.com>

* Update tests/test_generation_utils.py
Co-authored-by: default avatarPatrick von Platen <patrick.v.platen@gmail.com>

* Update tests/test_generation_utils.py
Co-authored-by: default avatarPatrick von Platen <patrick.v.platen@gmail.com>

* Update tests/test_generation_utils.py
Co-authored-by: default avatarPatrick von Platen <patrick.v.platen@gmail.com>

* Update tests/test_generation_utils.py

* fixing based on comments on PR

* took out the testing code that should but work fails without the beam search moditification ; style changes

* fixing comments issues

* docstrings for ConstraintListState

* typo in PhrsalConstraint docstring

* docstrings improvements

* finished adding what is sort of an opinionated implementation of disjunctive generation, but it revealed errors in inner beam search logic during testing.

* fixed bug found in constrained beam search that used beam_idx that were not global across all the batches

* disjunctive constraint working 100% correctly

* passing all tests

* Accidentally included mlruns

* Update src/transformers/generation_beam_constraints.py
Co-authored-by: default avatarPatrick von Platen <patrick.v.platen@gmail.com>

* Update src/transformers/generation_beam_constraints.py
Co-authored-by: default avatarPatrick von Platen <patrick.v.platen@gmail.com>

* complete overhaul of type complexities and other nits

* strict type checks in generate()

* fixing second round of feedback by narsil

* fixed failing generation test because of type check overhaul

* generation test fail fix

* fixing test fails
Co-authored-by: default avatarPatrick von Platen <patrick.v.platen@gmail.com>
      5c6f57ee
    • Javier de la Rosa's avatar
      Add missing support for Flax XLM-RoBERTa (#15900) · 01485cee
      Javier de la Rosa authored 
      

* Adding Flax XLM-RoBERTa

* Add Flax to __init__

* Adding doc and dummy objects

* Add tests

* Add Flax XLM-R models autodoc

* Fix tests

* Add Flask XLM-RoBERTa to TEST_FILES_WITH_NO_COMMON_TESTS

* Update src/transformers/models/xlm_roberta/modeling_flax_xlm_roberta.py
Co-authored-by: default avatarSuraj Patil <surajp815@gmail.com>

* Update tests/xlm_roberta/test_modeling_flax_xlm_roberta.py
Co-authored-by: default avatarSuraj Patil <surajp815@gmail.com>

* Update tests/xlm_roberta/test_modeling_flax_xlm_roberta.py
Co-authored-by: default avatarSuraj Patil <surajp815@gmail.com>

* Remove test on large Flask XLM-RoBERTa

* Add tokenizer to the test
Co-authored-by: default avatarSuraj Patil <surajp815@gmail.com>
      01485cee
    • Nicolas Patry's avatar
      Making MaskFormerForInstanceSegmentation. (#15934) · 89c7d9cf
      Nicolas Patry authored 
      Small adjustments.

Adding in type hint.

Last fix ?

Only include the default dict thing, not the pipelines.
      89c7d9cf