remove unnecessary include, add doc for the rest of params

Author: masahi

python/tvm/meta_schedule/schedule_rule/multi_level_tiling.py

@@ -93,7 +93,23 @@ class MultiLevelTilingWithIntrin(ScheduleRule):
     intrin_name : str
         The name of a tensor intrinsic, must be registerd via TensorIntrin.register(...) beforehand
 
-    The rest of parameters are the same as MultiLevelTiling above.
+    structure : str
+        The tiling structure. Recommended:
+        - 'SSRSRS' on CPU
+        - 'SSSRRSRS' on GPU
+    tile_bind : Optional[List[str]]
+        For each level of tiles, which thread axis it is bound to. Recommended:
+        - None on CPU
+        - [blockIdx.x, vthread.x, threadIdx.x] on GPU
+    max_innermost_factor : Optional[int]
+        The maximum size of the innermost factor. None means no limit
+    vector_load_lens : Optional[List[int]]
+        The length of vector lane in vectorized cooperative fetching.
+        None means disable vectorization
+    reuse_read : Optional[ReuseType]
+        Data reuse configuration for reading. None means no reuse.
+    reuse_write : Optional[ReuseType]
+        Data reuse configuration for writing. None means no reuse.
     """
 
     def __init__(

src/meta_schedule/schedule_rule/multi_level_tiling.cc

@@ -25,7 +25,6 @@
 #include <vector>
 
 #include "../utils.h"
-#include "tvm/meta_schedule/schedule_rule.h"
 
 namespace tvm {
 namespace tir {