Crates.io | treerite |
lib.rs | treerite |
version | 0.1.0 |
source | src |
created_at | 2021-01-13 00:39:07.999261 |
updated_at | 2021-01-13 00:39:07.999261 |
description | Rust binding for treelite runtime (https://github.com/dmlc/treelite) |
homepage | |
repository | https://github.com/dovahcrow/treerite |
max_upload_size | |
id | 341179 |
size | 17,206,474 |
This binding currently works for treelite 1.0.0rc1
Cargo.toml
.
treerite = { git = "https://github.com/dovahcrow/treerite" }
By default, the treerite library is static linked to your binary. If you'd like to use the dynamic lib,
set the dynamic feature of treerite
.
There's no documentation available yet. But you can take a look at the example folder and the tests folder.
The usage should be quite straight forward: you first load the Predictor
from a shared library of the model.
Then, load the data from Vec
or ndarray::Array2
into DMatrix
. Finally, you do prediction using
Predictor::predict_batch(dmatrix)
.