huggingface/optimum

Pull Requests Commits

Use transformers>=4.26.0 in setup.py (#723)

fxmarty committed 2 years ago

f4b4202a

fxmarty committed 2 years ago

d75b728a

Support generation config in ORTModel (#651)

fxmarty committed 2 years ago

9f9d997b

Release: v1.6.1

fxmarty committed 3 years ago

fdc08d78

Fix item access of some _TASKS_TO_AUTOMODELS (#642)

fxmarty committed 3 years ago

Verified 56651aa4

Revert breaking removal of EncoderOnnxConfig, DecoderOnnxConfig, _DecoderWithLMhead (#643)

fxmarty committed 3 years ago

Verified 60e25c9c

Update dev version (#639)

fxmarty committed 3 years ago

Verified 76b59adb

Allow kwargs in all generate_dummy_inputs() methods (#638)

fxmarty committed 3 years ago

Verified 433f477e

Add argument in the CLI to specify device to do the ONNX export on (#634)

fxmarty committed 3 years ago

Verified 874d2b25

Fix non-nesting bug in BetterTransformer integration (#637)

younesbelkada committed 3 years ago

Verified 46004525

Fixed GPU documentation for HF pipelines (#602)

smiraldr committed 3 years ago

Verified 5f00fee8

[`ONNX`] add `mobilenet` support (#633)

younesbelkada committed 3 years ago

Verified d3c28296

Handling ONNX models with external data (#586)

NouamaneTazi committed 3 years ago

Verified 6da9e1ad

Fixed FuseBiasInLinear by specifying device (#630)

IlyasMoutawwakil committed 3 years ago

Verified 0a02ea69

Improve `use_io_binding` default value for different execution providers (#604)

JingyaHuang committed 3 years ago

Verified 34e782e6

Support for custom input shapes in exporters onnx (#575)

fxmarty committed 3 years ago

Verified c3914e90

Use exporters in ORTTrainer (#546)

mht-sharma committed 3 years ago

Verified 178728b2

Remove diffusers dependency in onnxruntime code (#619)

fxmarty committed 3 years ago

Verified 36a01bfd

Add GPT-J normalized config (#623)

fxmarty committed 3 years ago

Verified 842eb914

Fix ORT GPU test (#624)

JingyaHuang committed 3 years ago

Verified 4c3b1c14

Set explictly the device index (#613)

JingyaHuang committed 3 years ago

Verified 356aa1d8

OnnxConfig now handle the export to encoder / decoder / decoder_with_past themselves (#590)

michaelbenayoun committed 3 years ago

Verified 10a21bb4

Use optimum exporter for current weight sharing test (#616)

JingyaHuang committed 3 years ago

Verified 83c941d6

More informative message at ONNX export (#609)

fxmarty committed 3 years ago

Verified a93df8d3

Documentation improvements (#598)

fxmarty committed 3 years ago

Verified 46d3ac0e

Fix documentation example (#603)

echarlaix committed 3 years ago

Verified e5e4f807

Merge ONNX decoder models (#587)

JingyaHuang committed 3 years ago

Verified 386e0ac0

Update dev version (#617)

fxmarty committed 3 years ago

Verified 7631364e

Add Optimum CLI backbone (#593)

fxmarty committed 3 years ago

Verified bc8f33ea

Add `encoder_last_hidden_state` as an output for encoder-decoder models (#601)

fxmarty committed 3 years ago

Verified 09afba1e

Older