Skip to content

Scalable, Portable and Distributed Gradient Boosting-



github pypi

How to install-

pip install --extra-index-url inaccel-xgboost

FPGA Platforms-

Get the available accelerators for your platform.

xilinx dynamic-shell

inaccel bitstream install

xilinx xdma_201830.2

inaccel bitstream install


Get Started with XGBoost-

This is a quick start tutorial for you to quickly try out XGBoost on the demo dataset on a classification task.

import xgboost as xgb

from sklearn.datasets import fetch_openml
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelEncoder
from sklearn.preprocessing import Normalizer
from timeit import default_timer as timestamp

X, y = fetch_openml('SVHN', return_X_y=True)

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.35)

features = Normalizer()
X_train = features.fit_transform(X_train)
X_test = features.transform(X_test)

label = LabelEncoder()
y_train = label.fit_transform(y_train)
y_test = label.transform(y_test)

params = {
    'alpha': 0.0,
    'eta': 0.3,
    'max_depth': 10,
    'num_class': len(label.classes_),
    'objective': 'multi:softmax',
    'subsample': 1.0,
    'tree_method': 'fpga_exact'

dtrain = xgb.DMatrix(X_train, y_train)
dtest = xgb.DMatrix(X_test, y_test)

begin = timestamp()
model = xgb.train(params, dtrain, 10)
end = timestamp()

print('time=%.3f' % (end - begin))

predictions = model.predict(dtest)

print('accuracy=%.3f' % accuracy_score(y_test, predictions))

Last update: May 15, 2022