(tvm) branch main updated: [DOC] Fix the loop length in a loop tiling example (#18753)

[tl...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

tlopex pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git


The following commit(s) were added to refs/heads/main by this push:
     new 4578a61bec [DOC] Fix the loop length in a loop tiling example (#18753)
4578a61bec is described below

commit 4578a61bec4f93ebeb6fdd0612376c4f46905dd3
Author: Ruslan Baratov <x...@ruslo.dev>
AuthorDate: Wed Feb 11 21:52:16 2026 +0800

    [DOC] Fix the loop length in a loop tiling example (#18753)
    
    According to the code, the loop is split by 16x8, not by 32x4
---
 docs/deep_dive/tensor_ir/tutorials/tir_transformation.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/deep_dive/tensor_ir/tutorials/tir_transformation.py b/docs/deep_dive/tensor_ir/tutorials/tir_transformation.py
index 724c47ed6c..26bc5e7a0d 100644
--- a/docs/deep_dive/tensor_ir/tutorials/tir_transformation.py
+++ b/docs/deep_dive/tensor_ir/tutorials/tir_transformation.py
@@ -109,7 +109,7 @@ i, j, k = sch.get_loops(block_Y)
 ######################################################################
 # We now proceed to execute the transformations. The initial modification involves
 # splitting loop ``j`` into two separate loops, with the inner loop possessing a
-# length of 4. It is crucial to understand that the transformation process is procedural;
+# length of 8. It is crucial to understand that the transformation process is procedural;
 # thus, inadvertent execution of the block twice will yield an error stating the
 # non-existence of variable ``j``.
 
@@ -122,7 +122,7 @@ sch.mod.show()
 
 ######################################################################
 # Following the initial transformation phase, two supplementary loops, ``j_0`` and ``j_1``,
-# have been generated with respective ranges of 32 and 4. The subsequent
+# have been generated with respective ranges of 16 and 8. The subsequent
 # action involves reordering these two loops.
 
 sch.reorder(j0, k, j1)