[IR] Fix assign_value_op bug (#56082)

* fix bug

* fix bug

paddle/fluid/ir/dialect/pd_op.yaml

+- name: assign_value
+  inputs: []
+  attrs:
+  - {typename: 'int[]', name: shape}
+  - {typename: DataType, name: dtype}
+  - {typename: 'Scalar[]', name: values, data_type: 'std::vector<Scalar>'}
+  - {typename: Place, name: place, default_value: '{}'}
+  outputs:
+  - {typename: Tensor, name: out, optional: false, intermediate: false}
+  no_need_buffer: null
+  data_transform: null
+  infer_meta:
+    func: AssignValueInferMeta
+    param: [shape, dtype]
+  kernel:
+    func: [assign_value]
+    param: [shape, dtype, values]
+    backend:
+      ordered: true
+      candidates: [place]
+    layout: null
+    data_type:
+      ordered: false
+      candidates: [dtype]
+      to_complex_flag: [false]
+    dispatch: {assign_value: null}
+    force_backend: null
+  inplace: null
+  view: null
+  backward: null
+
 - name: feed
   inputs: []
   attrs:

paddle/fluid/ir_adaptor/translator/op_translator.cc

@@ -761,11 +761,11 @@ struct IncrementOpTranscriber : public OpTranscriber {
 // python/paddle/tensor/creation.py::assign(x, output)
 struct AssignValueOpTranscriber : public OpTranscriber {
   ir::OpInfo LoopkUpOpInfo(ir::IrContext* ctx, const OpDesc& op_desc) override {
-    std::string target_op_name = "pd.assign_value_";
+    std::string target_op_name = "pd.assign_value";
     const auto& op_info = ctx->GetRegisteredOpInfo(target_op_name);
     if (!op_info) {
       IR_THROW(
-          "Op assign_value should have corresponding OpInfo pd.assign_value_");
+          "Op assign_value should have corresponding OpInfo pd.assign_value");
     }
 
     return op_info;
@@ -836,20 +836,7 @@ struct AssignValueOpTranscriber : public OpTranscriber {
 
     VLOG(10) << "[op assign_value] attribute translation done";
 
-    std::vector<int> src_shape =
-        paddle::get<std::vector<int>>(op_desc.GetAttr("shape"));
-    std::vector<int64_t> target_shape(src_shape.begin(), src_shape.end());
-
-    ir::Builder builder(ctx, program->block());
-    dialect::FullOp full_op = builder.Build<dialect::FullOp>(
-        target_shape,
-        0.0f,
-        attr_dtype.dyn_cast<dialect::DataTypeAttribute>().data(),
-        phi::CPUPlace());
-
-    std::vector<ir::OpResult> op_inputs = {full_op->result(0)};
-
-    VLOG(10) << "[op assign_value] insert a full op to get input";
+    std::vector<ir::OpResult> op_inputs = {};
 
     OpOutputMapping arg_to_idx;
     OpOutputTypeList op_output_types;