diff --git a/python/tvm/autotvm/graph_tuner/base_graph_tuner.py b/python/tvm/autotvm/graph_tuner/base_graph_tuner.py
index 53d91d2b878c2..90fdcd2d8d51f 100644
--- a/python/tvm/autotvm/graph_tuner/base_graph_tuner.py
+++ b/python/tvm/autotvm/graph_tuner/base_graph_tuner.py
@@ -54,7 +54,6 @@ def get_infer_layout(task_name):
 @autotvm.register_customized_task("layout_transform")
 def layout_transform(*args):
     """Autotvm layout transform template."""
-    args = deserialize_args(args)
     cfg = get_config()
     cfg.add_flop(-1)
     data = args[0]
diff --git a/python/tvm/relay/op/strategy/arm_cpu.py b/python/tvm/relay/op/strategy/arm_cpu.py
index a2468cc12c4f4..a7ba234dceb67 100644
--- a/python/tvm/relay/op/strategy/arm_cpu.py
+++ b/python/tvm/relay/op/strategy/arm_cpu.py
@@ -23,11 +23,11 @@
 from .. import op as _op
 
 @schedule_injective.register("arm_cpu")
-def schedule_injective(_, outs, target):
+def schedule_injective_arm_cpu(_, outs, target):
     with target:
         return topi.arm_cpu.schedule_injective(outs)
 
 @schedule_concatenate.register("arm_cpu")
-def schedule_concatenate(_, outs, target):
+def schedule_concatenate_arm_cpu(_, outs, target):
     with target:
         return topi.arm_cpu.schedule_concatenate(outs)
diff --git a/python/tvm/relay/op/strategy/cuda.py b/python/tvm/relay/op/strategy/cuda.py
index 97546567c0d4e..430647becc2e5 100644
--- a/python/tvm/relay/op/strategy/cuda.py
+++ b/python/tvm/relay/op/strategy/cuda.py
@@ -24,65 +24,65 @@
 from ....schedule import SpecializedCondition
 
 @schedule_injective.register(["cuda", "gpu"])
-def schedule_injective(attrs, outs, target):
+def schedule_injective_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_injective(outs)
 
 @schedule_reduce.register(["cuda", "gpu"])
-def schedule_reduce(attrs, outs, target):
+def schedule_reduce_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_reduce(outs)
 
 @schedule_concatenate.register(["cuda", "gpu"])
-def schedule_concatenate(attrs, outs, target):
+def schedule_concatenate_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_injective(outs)
 
 @schedule_pool.register(["cuda", "gpu"])
-def schedule_pool(attrs, outs, target):
+def schedule_pool_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_pool(outs, attrs.layout)
 
 @schedule_pool_grad.register(["cuda", "gpu"])
-def schedule_pool_grad(attrs, outs, target):
+def schedule_pool_grad_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_pool_grad(outs)
 
 @schedule_adaptive_pool.register(["cuda", "gpu"])
-def schedule_adaptive_pool(attrs, outs, target):
+def schedule_adaptive_pool_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_adaptive_pool(outs)
 
 @schedule_softmax.register(["cuda", "gpu"])
-def schedule_softmax(attrs, outs, target):
+def schedule_softmax_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_softmax(outs)
 
 @schedule_lrn.register(["cuda", "gpu"])
-def schedule_lrn(attrs, outs, target):
+def schedule_lrn_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_lrn(outs)
 
 @schedule_l2_normalize.register(["cuda", "gpu"])
-def schedule_l2_normalize(attrs, outs, target):
+def schedule_l2_normalize_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_l2_normalize(outs)
 
 @deformable_conv2d_strategy.register(["cuda", "gpu"])
-def deformable_conv2d_strategy(attrs, inputs, out_type, target):
+def deformable_conv2d_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_deformable_conv2d(topi.cuda.deformable_conv2d_nchw),
                            wrap_topi_schedule(topi.cuda.schedule_deformable_conv2d_nchw))
     return strategy
 
 @conv3d_strategy.register(["cuda", "gpu"])
-def conv3d_strategy(attrs, inputs, out_type, target):
+def conv3d_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     layout = attrs.data_layout
     assert layout in ["NCDHW", "NDHWC"], "Not support this layout {} yet".format(layout)
     if layout == "NCDHW":
         strategy.add_implement(wrap_compute_conv3d(topi.cuda.conv3d_ncdhw),
-                               wrap_topi_schedule(topi.cuda.schedule_conv3d_ncdhw),
+                               _reg._wrap_topi_schedule(topi.cuda.schedule_conv3d_ncdhw),
                                10)
     else: # layout == "NDHWC":
         strategy.add_implement(wrap_compute_conv3d(topi.cuda.conv3d_ndhwc),
@@ -95,7 +95,7 @@ def conv3d_strategy(attrs, inputs, out_type, target):
     return strategy
 
 @conv1d_transpose_strategy.register(["cuda", "gpu"])
-def conv1d_transpose_strategy(attrs, inputs, out_type, target):
+def conv1d_transpose_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     layout = attrs.data_layout
     dilation = get_const_tuple(attrs.dilation)
@@ -108,7 +108,7 @@ def conv1d_transpose_strategy(attrs, inputs, out_type, target):
     return strategy
 
 @dense_strategy.register(["cuda", "gpu"])
-def dense_strategy(attrs, inputs, out_type, target):
+def dense_strategy_cuda(attrs, inputs, out_type, target):
     # Todo(@icemelon9): update dense strategy
     strategy = _op.OpStrategy()
     if out_type.dtype == "int8":
@@ -127,8 +127,8 @@ def dense_strategy(attrs, inputs, out_type, target):
     return strategy
 
 @batch_matmul_strategy.register(["cuda", "gpu"])
-def batch_matmul_strategy(attrs, inputs, out_type, target):
-    strategy =_op.OpStrategy()
+def batch_matmul_strategy_cuda(attrs, inputs, out_type, target):
+    strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_batch_matmul(topi.nn.batch_matmul),
                            wrap_topi_schedule(topi.cuda.schedule_batch_matmul),
                            10)
@@ -139,57 +139,57 @@ def batch_matmul_strategy(attrs, inputs, out_type, target):
     return strategy
 
 @argsort_strategy.register(["cuda", "gpu"])
-def argsort_strategy(attrs, inputs, out_type, target):
+def argsort_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_argsort(topi.cuda.argsort_gpu),
                            wrap_topi_schedule(topi.cuda.schedule_argsort))
     return strategy
 
 @topk_strategy.register(["cuda", "gpu"])
-def topk_strategy(attrs, inputs, out_type, target):
+def topk_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_topk(topi.cuda.topk_gpu),
                            wrap_topi_schedule(topi.cuda.schedule_topk))
     return strategy
 
 @schedule_multibox_prior.register(["cuda", "gpu"])
-def schedule_multibox_prior(attrs, outs, target):
+def schedule_multibox_prior_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_multibox_prior(outs)
 
 @schedule_multibox_transform_loc.register(["cuda", "gpu"])
-def schedule_multibox_transform_loc(attrs, outs, target):
+def schedule_multibox_transform_loc_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_multibox_transform_loc(outs)
 
 @get_valid_counts_strategy.register(["cuda", "gpu"])
-def get_valid_counts_strategy(attrs, inputs, out_type, target):
+def get_valid_counts_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_get_valid_counts(topi.cuda.get_valid_counts),
                            wrap_topi_schedule(topi.cuda.schedule_get_valid_counts))
     return strategy
 
 @nms_strategy.register(["cuda", "gpu"])
-def nms_strategy(attrs, inputs, out_type, target):
+def nms_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_nms(topi.cuda.non_max_suppression),
                            wrap_topi_schedule(topi.cuda.schedule_nms))
     return strategy
 
 @roi_align_strategy.register(["cuda", "gpu"])
-def roi_align_strategy(attrs, inputs, out_type, target):
+def roi_align_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_roi_align(topi.vision.rcnn.roi_align_nchw),
                            wrap_topi_schedule(topi.cuda.schedule_roi_align))
     return strategy
 
 @schedule_roi_pool.register(["cuda", "gpu"])
-def schedule_roi_pool(attrs, outs, target):
+def schedule_roi_pool_cuda(attrs, outs, target):
     with target:
         return topi.cuda.schedule_roi_pool(outs)
 
 @proposal_strategy.register(["cuda", "gpu"])
-def proposal_strategy(attrs, inputs, out_type, target):
+def proposal_strategy_cuda(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_proposal(topi.cuda.proposal),
                            wrap_topi_schedule(topi.cuda.schedule_proposal))
diff --git a/python/tvm/relay/op/strategy/hls.py b/python/tvm/relay/op/strategy/hls.py
index fa78c7abe3b62..c943ad43aeac1 100644
--- a/python/tvm/relay/op/strategy/hls.py
+++ b/python/tvm/relay/op/strategy/hls.py
@@ -23,31 +23,31 @@
 from .. import op as _op
 
 @schedule_injective.register("hls")
-def schedule_injective(attrs, outs, target):
+def schedule_injective_hls(attrs, outs, target):
     with target:
         return topi.hls.schedule_injective(outs)
 
 @schedule_reduce.register("hls")
-def schedule_reduce(attrs, outs, target):
+def schedule_reduce_hls(attrs, outs, target):
     with target:
         return topi.hls.schedule_reduce(outs)
 
 @schedule_concatenate.register("hls")
-def schedule_concatenate(attrs, outs, target):
+def schedule_concatenate_hls(attrs, outs, target):
     with target:
         return topi.hls.schedule_injective(outs)
 
 @schedule_pool.register("hls")
-def schedule_pool(attrs, outs, target):
+def schedule_pool_hls(attrs, outs, target):
     with target:
         return topi.hls.schedule_pool(outs, attrs.layout)
 
 @schedule_adaptive_pool.register("hls")
-def schedule_adaptive_pool(attrs, outs, target):
+def schedule_adaptive_pool_hls(attrs, outs, target):
     with target:
         return topi.hls.schedule_adaptive_pool(outs)
 
 @schedule_softmax.register("hls")
-def schedule_softmax(attrs, outs, target):
+def schedule_softmax_hls(attrs, outs, target):
     with target:
         return topi.hls.schedule_softmax(outs)
\ No newline at end of file
diff --git a/python/tvm/relay/op/strategy/opengl.py b/python/tvm/relay/op/strategy/opengl.py
index 726bd4fc2d4e8..1e2c521733387 100644
--- a/python/tvm/relay/op/strategy/opengl.py
+++ b/python/tvm/relay/op/strategy/opengl.py
@@ -23,26 +23,26 @@
 from .. import op as _op
 
 @schedule_injective.register("opengl")
-def schedule_injective(attrs, outs, target):
+def schedule_injective_opengl(attrs, outs, target):
     with target:
         return topi.opengl.schedule_injective(outs)
 
 @schedule_concatenate.register("opengl")
-def schedule_injective(attrs, outs, target):
+def schedule_concatenate_opengl(attrs, outs, target):
     with target:
         return topi.opengl.schedule_injective(outs)
 
 @schedule_pool.register("opengl")
-def schedule_pool(attrs, outs, target):
+def schedule_pool_opengl(attrs, outs, target):
     with target:
         return topi.opengl.schedule_pool(outs, attrs.layout)
 
 @schedule_adaptive_pool.register("opengl")
-def schedule_adaptive_pool(attrs, outs, target):
+def schedule_adaptive_pool_opengl(attrs, outs, target):
     with target:
         return topi.opengl.schedule_adaptive_pool(outs)
 
 @schedule_softmax.register("opengl")
-def schedule_softmax(attrs, outs, target):
+def schedule_softmax_opengl(attrs, outs, target):
     with target:
         return topi.opengl.schedule_softmax(outs)
\ No newline at end of file
diff --git a/python/tvm/relay/op/strategy/rocm.py b/python/tvm/relay/op/strategy/rocm.py
index c878de70ad5dd..8901b42485164 100644
--- a/python/tvm/relay/op/strategy/rocm.py
+++ b/python/tvm/relay/op/strategy/rocm.py
@@ -23,11 +23,11 @@
 from .. import op as _op
 
 @schedule_lrn.register("rocm")
-def schedule_lrn(attrs, outs, target):
+def schedule_lrn_rocm(attrs, outs, target):
     with target:
         return topi.rocm.schedule_lrn(outs)
 
 @schedule_l2_normalize.register("rocm")
-def schedule_l2_normalize(attrs, outs, target):
+def schedule_l2_normalize_rocm(attrs, outs, target):
     with target:
         return topi.rocm.schedule_l2_normalize(outs)
diff --git a/python/tvm/relay/op/strategy/x86.py b/python/tvm/relay/op/strategy/x86.py
index f1e75e63b7a9a..717194400e3eb 100644
--- a/python/tvm/relay/op/strategy/x86.py
+++ b/python/tvm/relay/op/strategy/x86.py
@@ -24,37 +24,37 @@
 from ....schedule import SpecializedCondition
 
 @schedule_injective.register("cpu")
-def schedule_injective(attrs, outs, target):
+def schedule_injective_cpu(attrs, outs, target):
     with target:
         return topi.x86.schedule_injective(outs)
 
 @schedule_reduce.register("cpu")
-def schedule_reduce(attrs, outs, target):
+def schedule_reduce_cpu(attrs, outs, target):
     with target:
         return topi.x86.schedule_reduce(outs)
 
 @schedule_concatenate.register("cpu")
-def schedule_concatenate(attrs, outs, target):
+def schedule_concatenate_cpu(attrs, outs, target):
     with target:
         return topi.x86.schedule_concatenate(outs)
 
 @schedule_pool.register("cpu")
-def schedule_pool(attrs, outs, target):
+def schedule_pool_cpu(attrs, outs, target):
     with target:
         return topi.x86.schedule_pool(outs, attrs.layout)
 
 @schedule_adaptive_pool.register("cpu")
-def schedule_adaptive_pool(attrs, outs, target):
+def schedule_adaptive_pool_cpu(attrs, outs, target):
     with target:
         return topi.x86.schedule_adaptive_pool(outs)
 
 @schedule_softmax.register("cpu")
-def schedule_softmax(attrs, outs, target):
+def schedule_softmax_cpu(attrs, outs, target):
     with target:
         return topi.x86.schedule_softmax(outs)
 
 @conv2d_strategy.register("cpu")
-def conv2d_strategy(attrs, inputs, out_type, target):
+def conv2d_strategy_cpu(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     layout = attrs.data_layout
     dtype = out_type.dtype
@@ -68,7 +68,7 @@ def conv2d_strategy(attrs, inputs, out_type, target):
     return strategy
 
 @conv2d_NCHWc_strategy.register("cpu")
-def conv2d_NCHWc_strategy(attrs, inputs, out_type, target):
+def conv2d_NCHWc_strategy_cpu(attrs, inputs, out_type, target):
     print('inside x86 conv2d_NCHWc_strategy')
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_conv2d_NCHWc(topi.x86.conv2d_NCHWc),
@@ -76,7 +76,7 @@ def conv2d_NCHWc_strategy(attrs, inputs, out_type, target):
     return strategy
 
 @dense_strategy.register("cpu")
-def dense_strategy(attrs, inputs, out_type, target):
+def dense_strategy_cpu(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     m, k = inputs[0].shape
     strategy.add_implement(wrap_compute_dense(topi.x86.dense_nopack),
@@ -92,7 +92,7 @@ def dense_strategy(attrs, inputs, out_type, target):
     return strategy
 
 @batch_matmul_strategy.register("cpu")
-def batch_matmul_strategy(attrs, inputs, out_type, target):
+def batch_matmul_strategy_cpu(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_batch_matmul(topi.x86.batch_matmul),
                            wrap_topi_schedule(topi.x86.schedule_batch_matmul),
@@ -104,12 +104,12 @@ def batch_matmul_strategy(attrs, inputs, out_type, target):
     return strategy
 
 @schedule_sparse_dense.register("cpu")
-def schedule_sparse_dense(attrs, outs, target):
+def schedule_sparse_dense_cpu(attrs, outs, target):
     with target:
         return topi.x86.schedule_sparse_dense(outs)
 
 @roi_align_strategy.register("cpu")
-def roi_align_strategy(attrs, inputs, out_type, target):
+def roi_align_strategy_cpu(attrs, inputs, out_type, target):
     strategy = _op.OpStrategy()
     strategy.add_implement(wrap_compute_roi_align(topi.x86.roi_align_nchw),
                            wrap_topi_schedule(topi.generic.schedule_roi_align))