diff --git a/deepQlearn_0/src/deepQlearning/learn_to_drive.c b/deepQlearn_0/src/deepQlearning/learn_to_drive.c
index 7ddcaae..2681db6 100644
--- a/deepQlearn_0/src/deepQlearning/learn_to_drive.c
+++ b/deepQlearn_0/src/deepQlearning/learn_to_drive.c
@@ -74,13 +74,13 @@ struct networks_qlearning * create_network_qlearning(
   setup_networks_alloutputs_config_TYPE_FLOAT(&(qnets->best_net), config, false, minR, maxR, randomRange);  
   copy_weight_in_networks_from_main_to_best(qnets);
 
-  setup_all_layers_functions_TYPE_FLOAT(qnets->main_net, tensorContractnProdThread_TYPE_FLOAT, tensorProdThread_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
+  setup_all_layers_functions_TYPE_FLOAT(qnets->main_net, tensorContractnProdThreadOpt0_TYPE_FLOAT, tensorProdThread_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
   //setup_all_layers_functions_TYPE_FLOAT(qnets->main_net, tensorContractnProdTHR_TYPE_FLOAT, tensorProdTHR_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
   setup_all_layers_params_TYPE_FLOAT(qnets->main_net, nb_prod_thread, nb_calc_thread, learning_rate);
-  setup_all_layers_functions_TYPE_FLOAT(qnets->target_net, tensorContractnProdThread_TYPE_FLOAT, tensorProdThread_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
+  setup_all_layers_functions_TYPE_FLOAT(qnets->target_net, tensorContractnProdThreadOpt0_TYPE_FLOAT, tensorProdThread_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
   //setup_all_layers_functions_TYPE_FLOAT(qnets->target_net, tensorContractnProdTHR_TYPE_FLOAT, tensorProdTHR_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
   setup_all_layers_params_TYPE_FLOAT(qnets->target_net, nb_prod_thread, nb_calc_thread, learning_rate);
-  setup_all_layers_functions_TYPE_FLOAT(qnets->best_net, tensorContractnProdThread_TYPE_FLOAT, tensorProdThread_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
+  setup_all_layers_functions_TYPE_FLOAT(qnets->best_net, tensorContractnProdThreadOpt0_TYPE_FLOAT, tensorProdThread_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
   //setup_all_layers_functions_TYPE_FLOAT(qnets->best_net, tensorContractnProdTHR_TYPE_FLOAT, tensorProdTHR_TYPE_FLOAT, D_L2, L2, reLU, d_reLU);
   setup_all_layers_params_TYPE_FLOAT(qnets->best_net, nb_prod_thread, nb_calc_thread, learning_rate);