NielsRogge/transformers

Pull Requests Commits

Allow training to resume even if RNG states are not properly loaded (#14994)

sgugger committed 4 years ago

Verified e68c3756

Enabling `tokenizers` upgrade. (#14941)

Narsil committed 4 years ago

Verified 08cb5718

Adding `num_return_sequences` support for text2text generation. (#14988)

Narsil committed 4 years ago

Verified f8a989cf

[Generate] correct encoder_outputs are passed without attention_mask (#14980)

patrickvonplaten committed 4 years ago

Verified c043ce6c

[AutoProcessor] Correct AutoProcessor and automatically add processor… (#14881)

patrickvonplaten committed 4 years ago

Verified a1392883

Fixing a pathological case for slow tokenizers (#14981)

Narsil committed 4 years ago

Verified d7d60df0

remove absl workaround as it's no longer needed (#14909)

stas00 committed 4 years ago

Verified d1ba56d8

refactor: replace `assert` with `ValueError` (#14970)

jaketae committed 4 years ago

Verified 04cddaf4

[Wav2Vec2] Rename model's feature extractor to feature encoder (#14959)

patrickvonplaten committed 4 years ago

Verified 600496fa

[Tests] Speed up tokenizer tests (#14964)

patrickvonplaten committed 4 years ago

Verified 1bfa3477

Update README.md (#14965)

patrickvonplaten committed 4 years ago

Verified f80775df

[WavLM] give model for precision (#14958)

patrickvonplaten committed 4 years ago

Verified 1e847b40

Add Speech Seq2Seq Training script (#14792)

patrickvonplaten committed 4 years ago

Verified 1c121916

[doc] :class: hunt (#14955)

stas00 committed 4 years ago

Verified 10fd4fa1

sgugger committed 4 years ago

2c5597f6

Doc styler examples (#14953)

sgugger committed 4 years ago

Verified b5e2b183

[doc] :obj: hunt (#14954)

stas00 committed 4 years ago

Verified e13f72fb

[doc] consistent True/False/None default format (#14951)

stas00 committed 4 years ago

Verified 133c5e40

Convert last rst file (#14952)

sgugger committed 4 years ago

Verified b2f50025

Doc styler v2 (#14950)

sgugger committed 4 years ago

Verified 87e6e4fe

Fix duplicate call to save_checkpoint when using deepspeed (#14946)

MihaiBalint committed 4 years ago

Verified c1138273

fix to issue #14833 in data_collator - consider no labels (#14930)

Ayal Klein committed 4 years ago

Verified 03885a3f

Add `ElectraForCausalLM` -> Enable Electra encoder-decoder model (#14729)

stancld committed 4 years ago

Verified 501307b5

ChunkPipeline (batch_size enabled on `zero-cls` and `qa` pipelines. (#14225)

Narsil committed 4 years ago

Verified b058490c

Fix Perceiver docs (#14917)

Sanster committed 4 years ago

Verified 705ca7f2

[WavLM] fix wavlm docs (#14910)

patrickvonplaten committed 4 years ago

Verified 11682990

[doc] install - add jax (#14912)

stas00 committed 4 years ago

Verified 41581066

Better logic for getting tokenizer config in AutoTokenizer (#14906)

sgugger committed 4 years ago

Verified 676643c6

Fix failing GPU trainer tests (#14903)

sgugger committed 4 years ago

Verified f566c6e3

[Generate] Remove attention_mask and integrate model_main_input_name (#14856)

patrickvonplaten committed 4 years ago

Verified fe4197ab

Newer Older