ONNX

Using the Open Neural Network Exchange (ONNX) format to load pre-trained weights and use them for prediction. This allows using models trained via frameworks like PyTorch, TensorFlow, Scikit-Learn etc in TurboML. Note: This model doesn't learn.

Parameters

• model_save_name(str) → The name used to save the ONNX model weights.

Example Usage

We can create an instance and deploy ONNX model like this.

import turboml as tb
tb.set_onnx_model("randomforest", onx.SerializeToString())
onnx_model = tb.ONNX(model_save_name = "randomforest")