huggingface/optimum

Pull Requests Commits

Release: v1.4.0

echarlaix committed 3 years ago

1b08fb5a

Adapt BaseConfig for v4.22 transformers release (#368)

echarlaix committed 3 years ago

Verified b1bef7c9

Add default clone url for making doc (#373)

regisss committed 3 years ago

Verified b0d93704

Add habana-deepspeed extra to setup.py (#370)

regisss committed 3 years ago

Verified bd54d547

Fix documentation (#369)

philschmid committed 3 years ago

Verified 815a7628

Fix filenames when empty suffix given (#363)

Pierre Snell committed 3 years ago

Verified cc5bb35a

Fix import (#360)

ierezell committed 3 years ago

Verified 661f4423

ORT optimizer refactorization (#294)

echarlaix committed 3 years ago

Verified fb7e303d

Do not force download by default in ORTModel (#356)

fxmarty committed 3 years ago

Verified 122a9d8d

Add ONNX Runtime inference for models with multiple choice classification head (#358)

echarlaix committed 3 years ago

Verified 58360c7d

Add documentation dispatch (#353)

philschmid committed 3 years ago

Verified e9033246

Fix bug where the PR doc is actually the main doc (#349)

regisss committed 3 years ago

Verified 26661dd6

Fix failing ORT Windows test (#352)

regisss committed 3 years ago

Verified 107f5784

Fix typo in variable name (#343)

mausch committed 3 years ago

Verified 41688286

Fix the way the linear modules are removed from the parent module (#342)

michaelbenayoun committed 3 years ago

Verified aed90206

Refactor ORTQuantizer (#270)

philschmid committed 3 years ago

Verified 69f2883d

Allow string device input for OrtModel.to(device) (#337)

jessecambon committed 3 years ago

Verified f69264f9

Add support of big-bird in config for ORTOptimizer (#329)

lewisbails committed 3 years ago

Verified 0c0c5ca5

Updated the operators_to_quantize doc string for the QuantizationConfig class (#296)

Shamik-07 committed 3 years ago

Verified 7086ba5d

Fix bug in docs/combine_docs.py when there are subsections (#335)

regisss committed 3 years ago

Verified 1a6864af

Reschedule ORTTrainer CI(#339)

JingyaHuang committed 3 years ago

Verified b91dc882

Update train docker to onnxruntime-training 1.12.0 (#328)

JingyaHuang committed 3 years ago

Verified 609ab61c

Add support of xlm-roberta in config for ORTOptimizer (#334)

jegork committed 3 years ago

Verified 7eb5b669

Add ORTModel support for custom tasks (#303)

JingyaHuang committed 3 years ago

Verified d3c0b754

Fix seq2seq onnx config (#317)

echarlaix committed 3 years ago

Verified 915c1f11

remove transformers main in benchmark test (#330)

fxmarty committed 3 years ago

Verified ee835054

OnnxRuntime tests refactor to make them faster (#315)

michaelbenayoun committed 3 years ago

Verified db845bd1

FuseBiasInLinear transformation (#253)

michaelbenayoun committed 3 years ago

Verified d03433ea

MergeLinears set proper device and dtype (#313)

michaelbenayoun committed 3 years ago

Verified 23cbf75b

Enable possibility to not use past_key_values (#241)

echarlaix committed 3 years ago

Verified 8d64d8aa

Older