fix grpah mode loop sink bug in auto parallel

mindspore/ccsrc/parallel/auto_parallel/graph_costmodel.h

@@ -34,8 +34,8 @@ namespace parallel {
 #define OPERATOR_TO_OPERATOR_CONNECTOR "-"
 #define DEFAULT_DEVICE_MEMORY_CAPACITY (1024.0 * 1024.0 * 1024.0 * 16.0)
 #define DEFAULT_COST_MODEL_ALPHA 1.0
-#define DEFAULT_COST_MODEL_BETA 65.0
-#define DEFAULT_COST_MODEL_GAMMA 0.02
+#define DEFAULT_COST_MODEL_BETA 260.0
+#define DEFAULT_COST_MODEL_GAMMA 0.001
 #define DEFAULT_COST_MODEL_SIMPLIFY_CALCULATION true
 #define DEFAULT_COST_MODEL_COMMUNI_THRESHOLD 2048.0
 #define DEFAULT_COST_MODEL_COMMUNI_CONST 3072.0

mindspore/ccsrc/parallel/step_parallel.cc

@@ -375,6 +375,10 @@ bool IsParallelCareNode(const CNodePtr& cnode) {
     MS_LOG(INFO) << "Parallel don't care node: " << prim->name();
     return false;
   }
+  // get_next is not in the forward graph, we need mark the get_next as the forward node
+  if (prim->name() == GET_NEXT) {
+    return true;
+  }
   if ((prim->name() == CAST)) {
     if ((!attrs.count(STRATEGY)) && (cnode->operator_info() == nullptr)) {
       return false;

mindspore/train/dataset_helper.py

@@ -88,7 +88,7 @@ class _DatasetIter:
         # times the batch dimension of tensors for run
         if _get_parallel_mode() in (ParallelMode.SEMI_AUTO_PARALLEL, ParallelMode.AUTO_PARALLEL):
             device_num = _get_device_num()
-            dataset_shapes = _to_full_shapes(dataset_shapes, device_num)
+            self.dataset_shapes = _to_full_shapes(dataset_shapes, device_num)
 
     def __iter__(self):
         self.ind = 0

tests/ut/python/parallel/test_auto_parallel_parameter_cast.py

@@ -80,9 +80,9 @@ def test_common_parameter():
 
     _executor.compile(net, x, y, z, w, phase='train')
     strategies = _executor._get_strategy(net)
-    expected_strategies = {'Default/network-Net/MatMul-op8': [[1, 1], [1, 8]],
-                           'Default/network-Net/MatMul-op9': [[1, 1], [1, 8]],
-                           'Default/network-Net/Cast-op10': [[1, 8]],
-                           'Default/network-Net/MatMul-op0': [[1, 1], [1, 8]],
-                           'Default/network-Net/Cast-op11': [[1, 8]]}
-    assert strategies == expected_strategies
+    expected_strategies = {'Default/network-Net/MatMul-op6': [[8, 1], [1, 1]],
+                           'Default/network-Net/MatMul-op8': [[8, 1], [1, 1]],
+                           'Default/network-Net/Cast-op7': [[1, 1]],
+                           'Default/network-Net/MatMul-op0': [[8, 1], [1, 1]],
+                           'Default/network-Net/Cast-op9': [[1, 1]]}
+    assert strategies == expected_strategies

tests/ut/python/parallel/test_auto_parallel_two_matmul.py

@@ -86,9 +86,9 @@ def test_two_matmul():
     costmodel_alpha = cost_model_context.get_cost_model_context("costmodel_alpha")
     assert costmodel_alpha == 1.0
     costmodel_beta = cost_model_context.get_cost_model_context("costmodel_beta")
-    assert costmodel_beta == 65.0
+    assert costmodel_beta == 260.0
     costmodel_gamma = cost_model_context.get_cost_model_context("costmodel_gamma")
-    assert costmodel_gamma == 0.02
+    assert costmodel_gamma == 0.001
     costmodel_communi_threshold = cost_model_context.get_cost_model_context("costmodel_communi_threshold")
     assert costmodel_communi_threshold == 2048.0
     costmodel_communi_const = cost_model_context.get_cost_model_context("costmodel_communi_const")
@@ -137,4 +137,5 @@ def test_two_matmul():
     strategies = _executor._get_strategy(net)
     expected_strategies = {'Default/network-Net/MatMul-op2': [[16, 1], [1, 1]],
                      'Default/network-Net/MatMul-op3': [[16, 1], [1, 1]]}
-    assert strategies == expected_strategies
+    assert strategies == expected_strategies
+