u2d

JaxGaussianProcesses · Oct 22, 2023 · 8bbce75 · 8bbce75
1 parent 6eca0ec
commit 8bbce75
Show file tree

Hide file tree

Showing 3 changed files with 3 additions and 18 deletions.
diff --git a/gpjax/__init__.py b/gpjax/__init__.py
@@ -22,14 +22,10 @@
 )
 from gpjax.citation import cite
 from gpjax.dataset import Dataset
-<<<<<<< HEAD
 from gpjax.fit import (
     fit,
     fit_scipy,
 )
-=======
-from gpjax.fit import fit
->>>>>>> main
 from gpjax.gps import (
     Prior,
     construct_posterior,
@@ -94,10 +90,7 @@
     "decision_making",
     "kernels",
     "fit",
-<<<<<<< HEAD
     "fit_scipy",
-=======
->>>>>>> main
     "Prior",
     "construct_posterior",
     "integrators",

diff --git a/gpjax/fit.py b/gpjax/fit.py
@@ -26,10 +26,7 @@
 from jax._src.random import _check_prng_key
 import jax.numpy as jnp
 import jax.random as jr
-<<<<<<< HEAD
 import jaxopt
-=======
->>>>>>> main
 import optax as ox
 
 from gpjax.base import Module

diff --git a/tests/test_fit.py b/tests/test_fit.py
@@ -30,6 +30,7 @@
 from gpjax.dataset import Dataset
 from gpjax.fit import (
     fit,
+    fit_scipy,
     get_batch,
 )
 from gpjax.gps import (
@@ -97,9 +98,8 @@ def step(self, model: LinearModel, train_data: Dataset) -> float:
     # Test stop_gradient on bias:
     assert trained_model.bias == 1.0
 
-<<<<<<< HEAD
     # Train with bfgs!
-    trained_model, hist = fit_bfgs(
+    trained_model, hist = fit_scipy(
         model=model,
         objective=loss,
         train_data=D,
@@ -118,8 +118,6 @@ def step(self, model: LinearModel, train_data: Dataset) -> float:
     # Test stop_gradient on bias:
     assert trained_model.bias == 1.0
 
-=======
->>>>>>> main
 
 @pytest.mark.parametrize("num_iters", [1, 5])
 @pytest.mark.parametrize("n_data", [1, 20])
@@ -161,9 +159,8 @@ def test_gaussian_process_regression(num_iters, n_data: int, verbose: bool) -> N
     # Ensure we reduce the loss
     assert mll(trained_model, D) < mll(posterior, D)
 
-<<<<<<< HEAD
     # Train with BFGS!
-    trained_model_bfgs, history_bfgs = fit_bfgs(
+    trained_model_bfgs, history_bfgs = fit_scipy(
         model=posterior,
         objective=mll,
         train_data=D,
@@ -180,8 +177,6 @@ def test_gaussian_process_regression(num_iters, n_data: int, verbose: bool) -> N
     # Ensure we reduce the loss
     assert mll(trained_model_bfgs, D) < mll(posterior, D)
 
-=======
->>>>>>> main
 
 @pytest.mark.parametrize("num_iters", [1, 5])
 @pytest.mark.parametrize("batch_size", [1, 20, 50])