diff --git a/test/test_fori_loop_simple_linear_model_test_code.py b/test/test_fori_loop_simple_linear_model_test_code.py
index f43a040fb2d..e57c1e3f013 100644
--- a/test/test_fori_loop_simple_linear_model_test_code.py
+++ b/test/test_fori_loop_simple_linear_model_test_code.py
@@ -78,8 +78,8 @@ def body_fun(l_in_i):
 # lower_, upper_, one_value_, add_res_, l_out_res_, weight_, final_one_= fori_loop(upper, lower, body_fun, one_value, init_val, l_in_0) # , placeholder_func, placeholder_input) # , linear_0, l_in_0)
 # lower_, upper_, one_value_, add_res_, l_out_res_, weight_, final_one_= fori_loop(lower, upper, body_fun, one_value, init_val, l_in_0) # , placeholder_func, placeholder_input) # , linear_0, l_in_0)
 # one_value_, upper_, lower_, add_res_x_, bias_, weight_, l_in_i_plus_1_, l_out_= fori_loop(upper, lower, body_fun, one_value, init_val, l_in_0) # , placeholder_func, placeholder_input) # , linear_0, l_in_0)
-weight_0 =linear_0.weight
-bias_0 = linear_0.bias
+# weight_0 =linear_0.weight
+# bias_0 = linear_0.bias
 # one_value_, lower_, upper_, add_res_x_, bias_, weight_, l_in_i_plus_1_, l_out_= fori_loop(one_value, lower, upper, body_fun, init_val, l_in_0, weight_0=weight_0, bias_0=bias_0) # , placeholder_func, placeholder_input) # , linear_0, l_in_0)
 one_value_, lower_, upper_, add_res_x_, bias_, weight_, l_in_i_plus_1_, l_out_= fori_loop(one_value, lower, upper, linear_0, init_val, l_in_0) #, weight_0=weight_0, bias_0=bias_0) # , placeholder_func, placeholder_input) # , linear_0, l_in_0)
 #one_value, [upper],[lower],x,          [bias],[new_weight], [l_in_i+1], l_out
diff --git a/torch_xla/experimental/fori_loop.py b/torch_xla/experimental/fori_loop.py
index f07f8882534..6baccd7253f 100644
--- a/torch_xla/experimental/fori_loop.py
+++ b/torch_xla/experimental/fori_loop.py
@@ -13,9 +13,11 @@
 
 
 # TODO(@manfei): delete one_value?
-def fori_loop(one_value, lower, upper, body_fun, init_val, *input_value, weight_0, bias_0):
+def fori_loop(one_value, lower, upper, body_fun, init_val, *input_value): #, weight_0, bias_0):
 
   device = xm.xla_device()
+  weight_0 =linear_0.weight
+  bias_0 = linear_0.bias
 
   # a = torch.tensor(1, dtype=torch.int32, device=device)
   # b = torch.tensor(1, dtype=torch.int32, device=device)