Remove duplicate call to objective function in strong wolfe line search in L-BFGS optimizer. (#72773)

svenslaggare · pytorchmergebot · commit f43165a75ff6 · 2022-02-15T15:33:13.000Z
Summary: With this change, the optimizer is almost twice as fast as before. As the result of the first call is never used, it looks like a copy paste error and therefore can be removed. In addition, this duplicate call is not present in the Python implementation. Pull Request resolved: #72773 Reviewed By: samdow Differential Revision: D34214312 Pulled By: albanD fbshipit-source-id: 4f4de08633c7236f3ccce8a2a74e56500003281b (cherry picked from commit 4a63f81)
diff --git a/torch/csrc/api/src/optim/lbfgs.cpp b/torch/csrc/api/src/optim/lbfgs.cpp
@@ -232,7 +232,6 @@ std::tuple<double, Tensor, double, int64_t> _strong_wolfe(const Function& obj_fu
     auto d_norm = val(d.abs().max());
     g = g.clone(at::MemoryFormat::Contiguous);
     // evaluate objective and gradient using initial step
-    auto obj_func_res = obj_func(x, t, d);
     // NOLINTNEXTLINE(cppcoreguidelines-init-variables)
     double f_new;
     Tensor g_new;
@@ -285,7 +284,6 @@ std::tuple<double, Tensor, double, int64_t> _strong_wolfe(const Function& obj_fu
       f_prev = f_new;
       g_prev = g_new.clone(at::MemoryFormat::Contiguous);
       gtd_prev = gtd_new;
-      obj_func_res = obj_func(x, t, d);
       std::tie(f_new, g_new) = obj_func(x, t, d);
       ls_func_evals += 1;
       gtd_new = g_new.dot(d);
@@ -335,9 +333,7 @@ std::tuple<double, Tensor, double, int64_t> _strong_wolfe(const Function& obj_fu
       }
 
       // Evaluate new point
-      obj_func_res = obj_func(x, t, d);
-      f_new = std::get<0>(obj_func_res);
-      g_new = std::get<1>(obj_func_res);
+      std::tie(f_new, g_new) = obj_func(x, t, d);
       ls_func_evals += 1;
       gtd_new = g_new.dot(d);
       ls_iter += 1;
