fix lint

Author: merrymercy

python/tvm/auto_scheduler/auto_schedule.py

@@ -159,6 +159,7 @@ def __init__(
             measure_callbacks,
         )
 
+
 def create_task(func, args, target, target_host=None, hardware_params=None):
     """Create a search task
 
@@ -184,6 +185,7 @@ def create_task(func, args, target, target_host=None, hardware_params=None):
     dag = ComputeDAG(workload_key)
     return SearchTask(dag, workload_key, target, target_host, hardware_params)
 
+
 def auto_schedule(task, search_policy=None, tuning_options=TuningOptions()):
     """Run auto scheduling search for a task
 

tutorials/auto_scheduler/tune_matmul_x86.py

@@ -41,25 +41,27 @@
 # The function should return the list of input/output tensors.
 # From these tensors, the auto-scheduler can get the whole computational graph.
 
+
 @auto_scheduler.register_workload
 def matmul_add(N, L, M, dtype):
-    A = te.placeholder((N, L), name='A', dtype=dtype)
-    B = te.placeholder((L, M), name='B', dtype=dtype)
-    C = te.placeholder((N, M), name='C', dtype=dtype)
+    A = te.placeholder((N, L), name="A", dtype=dtype)
+    B = te.placeholder((L, M), name="B", dtype=dtype)
+    C = te.placeholder((N, M), name="C", dtype=dtype)
 
-    k = te.reduce_axis((0, L), name='k')
-    matmul = te.compute((N, M), lambda i, j: te.sum(A[i, k] * B[k, j], axis=k), name='matmul')
-    out = te.compute((N, M), lambda i, j: matmul[i, j] + C[i, j], name='D')
+    k = te.reduce_axis((0, L), name="k")
+    matmul = te.compute((N, M), lambda i, j: te.sum(A[i, k] * B[k, j], axis=k), name="matmul")
+    out = te.compute((N, M), lambda i, j: matmul[i, j] + C[i, j], name="D")
 
     return [A, B, C, out]
 
+
 ######################################################################
 # Create the search task
 # ^^^^^^^^^^^^^^^^^^^^^^
-# We then create the a search task with N=L=M=128 and dtype='float32'
+# We then create the a search task with N=L=M=128 and dtype="float32"
 
 target = tvm.target.Target("llvm")
-task = auto_scheduler.create_task(matmul_add, (128, 128, 128, 'float32'), target)
+task = auto_scheduler.create_task(matmul_add, (128, 128, 128, "float32"), target)
 
 # inspect the computational graph
 print(task.compute_dag)
@@ -68,15 +70,16 @@ def matmul_add(N, L, M, dtype):
 # Next, we set parameters for the auto-scheduler.
 #
 # * `num_measure_trials` is the number of measurement trials we can use during the search.
-#   We only make 10 trials in this tutorial for a fast demonstration. In practice, 1000 is a 
+#   We only make 10 trials in this tutorial for a fast demonstration. In practice, 1000 is a
 #   good value for the search to converge. You can do more trials according to your time budget.
 # * In addition, we use `RecordToFile` to dump measurement records into a file `matmul.json`.
 #   The measurement records can be used to query the history best, resume the search,
 #   or do more analysis later.
 # * see :any:`auto_schedule.TuningOptions`: for more parameters
 
-tune_option = auto_scheduler.TuningOptions(num_measure_trials=100,
-                                           measure_callbacks=[auto_scheduler.RecordToFile('matmul.json')])
+tune_option = auto_scheduler.TuningOptions(
+    num_measure_trials=10, measure_callbacks=[auto_scheduler.RecordToFile('matmul.json')]
+)
 
 ######################################################################
 # Run the search
@@ -85,8 +88,7 @@ def matmul_add(N, L, M, dtype):
 # We can kick off the search and let the auto-scheduler do its magic.
 # After some measurement trials, it will return the best schedule it founds.
 
-sch, args = auto_scheduler.auto_schedule(task,
-                                         tuning_options=tune_option)
+sch, args = auto_scheduler.auto_schedule(task, tuning_options=tune_option)
 
 ######################################################################
 # We can lower schedule to see the IR after auto-scheduling.
@@ -113,19 +115,19 @@ def matmul_add(N, L, M, dtype):
 # Using the record file
 # ^^^^^^^^^^^^^^^^^^^^^
 # During the search, all measuremnt records is dumpped into the record
-# file 'matmul.json'. The measurement records can be used to resume the 
+# file "matmul.json". The measurement records can be used to resume the 
 # search, re-apply search results and other analysis.
 #
 # Here we show an example where we load the best schedule from a file,
 # print the equivalent python schedule API, and build the binary again.
 
-inp, res = auto_scheduler.load_best('matmul.json', task.workload_key)
+inp, res = auto_scheduler.load_best("matmul.json", task.workload_key)
 
 # Print equivalent python schedule API. This can be used for debugging and
-# learn the behavior of auto-scheduler.
+# learning the behavior of auto-scheduler.
 print(task.compute_dag.print_python_code_from_state(inp.state))
 
-# Rebuild the binary. This shows how you can apply the best schedule from a 
+# Rebuild the binary. This shows how you can apply the best schedule from a
 # log file without reruning the search again.
 sch, args = task.compute_dag.apply_steps_from_state(inp.state)
 func = tvm.build(sch, args)