1. 15 Dec, 2020 12 commits
  2. 14 Dec, 2020 5 commits
    • Julien Plu's avatar
      Fix T5 and BART for TF (#9063) · df3f4d2a
      Julien Plu authored
      * Fix T5 for graphe compilation+execution
      
      * Fix BART
      
      * Fix import
      
      * Fix naming
      
      * fix attribute name
      
      * Oops
      
      * fix import
      
      * fix tests
      
      * fix tests
      
      * Update test
      
      * Add mising import
      
      * Address Patrick's comments
      
      * Style
      
      * Address Patrick's comment
      df3f4d2a
    • Ahmed Elnaggar's avatar
      Add parallelization support for T5EncoderModel (#9082) · a9c8bff7
      Ahmed Elnaggar authored
      
      
      * add model parallelism to T5EncoderModel
      
      add model parallelism to T5EncoderModel
      
      * remove decoder from T5EncoderModel parallelize
      
      * uodate T5EncoderModel docs
      
      * Extend T5ModelTest for T5EncoderModel
      
      * fix T5Stask using range for get_device_map
      
      * fix style
      Co-authored-by: default avatarAhmed Elnaggar <elnaggar@rostlab.informatik.tu-muenchen.de>
      a9c8bff7
    • Navjot's avatar
      d6af344c
    • Patrick von Platen's avatar
      [RAG, Bart] Align RAG, Bart cache with T5 and other models of transformers (#9098) · fa1ddced
      Patrick von Platen authored
      * fix rag
      
      * fix slow test
      
      * fix past in bart
      fa1ddced
    • Julien Plu's avatar
      Fix embeddings resizing in TF models (#8657) · 51d9c569
      Julien Plu authored
      * Resize the biases in same time than the embeddings
      
      * Trigger CI
      
      * Biases are not reset anymore
      
      * Remove get_output_embeddings + better LM model detection in generation utils
      
      * Apply style
      
      * First test on BERT
      
      * Update docstring + new name
      
      * Apply the new resizing logic to all the models
      
      * fix tests
      
      * Apply style
      
      * Update the template
      
      * Fix naming
      
      * Fix naming
      
      * Apply style
      
      * Apply style
      
      * Remove unused import
      
      * Revert get_output_embeddings
      
      * Trigger CI
      
      * Update num parameters
      
      * Restore get_output_embeddings in TFPretrainedModel and add comments
      
      * Style
      
      * Add decoder resizing
      
      * Style
      
      * Fix tests
      
      * Separate bias and decoder resize
      
      * Fix tests
      
      * Fix tests
      
      * Apply style
      
      * Add bias resizing in MPNet
      
      * Trigger CI
      
      * Apply style
      51d9c569
  3. 11 Dec, 2020 3 commits
  4. 10 Dec, 2020 5 commits
  5. 09 Dec, 2020 6 commits
  6. 08 Dec, 2020 6 commits
    • Lysandre Debut's avatar
      Templates overhaul 1 (#8993) · 67ff1c31
      Lysandre Debut authored
      67ff1c31
    • guillaume-be's avatar
      Removed unused `encoder_hidden_states` and `encoder_attention_mask` (#8972) · 7809eb82
      guillaume-be authored
      * Removed unused `encoder_hidden_states` and `encoder_attention_mask` from MobileBert
      
      * Removed decoder tests for MobileBert
      
      * Removed now unnecessary import
      7809eb82
    • Lysandre Debut's avatar
    • Sylvain Gugger's avatar
      Make `ModelOutput` pickle-able (#8989) · 04c446f7
      Sylvain Gugger authored
      04c446f7
    • Julien Plu's avatar
      Optional layers (#8961) · bf7f79cd
      Julien Plu authored
      * Apply on BERT and ALBERT
      
      * Update TF Bart
      
      * Add input processing to TF BART
      
      * Add input processing for TF CTRL
      
      * Add input processing to TF Distilbert
      
      * Add input processing to TF DPR
      
      * Add input processing to TF Electra
      
      * Add deprecated arguments
      
      * Add input processing to TF XLM
      
      * remove unused imports
      
      * Add input processing to TF Funnel
      
      * Add input processing to TF GPT2
      
      * Add input processing to TF Longformer
      
      * Add input processing to TF Lxmert
      
      * Apply style
      
      * Add input processing to TF Mobilebert
      
      * Add input processing to TF GPT
      
      * Add input processing to TF Roberta
      
      * Add input processing to TF T5
      
      * Add input processing to TF TransfoXL
      
      * Apply style
      
      * Rebase on master
      
      * Fix wrong model name
      
      * Fix BART
      
      * Apply style
      
      * Put the deprecated warnings in the input processing function
      
      * Remove the unused imports
      
      * Raise an error when len(kwargs)>0
      
      * test ModelOutput instead of TFBaseModelOutput
      
      * Address Patrick's comments
      
      * Address Patrick's comments
      
      * Add boolean processing for the inputs
      
      * Take into account the optional layers
      
      * Add missing/unexpected weights in the other models
      
      * Apply style
      
      * rename parameters
      
      * Apply style
      
      * Remove useless
      
      * Remove useless
      
      * Remove useless
      
      * Update num parameters
      
      * Fix tests
      
      * Address Patrick's comment
      
      * Remove useless attribute
      bf7f79cd
    • Stas Bekman's avatar
      [training] SAVE_STATE_WARNING was removed in pytorch (#8979) · 9d7d0005
      Stas Bekman authored
      * [training] SAVE_STATE_WARNING was removed in pytorch
      
      FYI `SAVE_STATE_WARNING` has been removed 3 days ago: pytorch/pytorch#46813
      
      Fixes: #8232
      
      @sgugger
      
      * style, but add () to prevent autoformatters from botching it
      
      * switch to try/except
      
      * cleanup
      9d7d0005
  7. 07 Dec, 2020 3 commits