Skip to content

Latest commit

 

History

History
8 lines (8 loc) · 278 Bytes

File metadata and controls

8 lines (8 loc) · 278 Bytes

Use C++ interface to use TurboTransformers

  1. prepare your model in format of *.npz bash $WORKSPACE/tools/convert_huggingface_bert_pytorch_to_npz.py bert-base-uncased bert.npz move bert.npz to $YOUR_BUILD_DIR/example/cpp/models
  2. run our example
./bert_model_example