You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by ma...@apache.org on 2020/09/29 10:00:32 UTC

[incubator-tvm] branch master updated: disable stacked bidir test (#6585)

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

masahi pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git


The following commit(s) were added to refs/heads/master by this push:
     new 27abfad  disable stacked bidir test (#6585)
27abfad is described below

commit 27abfadc55e79e3d40b7de7d4b87eb87a19b7b97
Author: masahi <ma...@gmail.com>
AuthorDate: Tue Sep 29 19:00:14 2020 +0900

    disable stacked bidir test (#6585)
    
    Co-authored-by: masa <ma...@pop-os.localdomain>
---
 tests/python/frontend/pytorch/test_lstm.py | 22 +++++++++++++++-------
 1 file changed, 15 insertions(+), 7 deletions(-)

diff --git a/tests/python/frontend/pytorch/test_lstm.py b/tests/python/frontend/pytorch/test_lstm.py
index 27dbec3..4d7a406 100644
--- a/tests/python/frontend/pytorch/test_lstm.py
+++ b/tests/python/frontend/pytorch/test_lstm.py
@@ -317,17 +317,24 @@ def test_custom_lstm():
     ]
 
     models = [
-        (lstm(input_size, hidden_size).eval(), states[0], input_shapes),
-        (stacked_lstm(input_size, hidden_size, num_layers).eval(), states, input_shapes_stacked),
-        (bidir_lstm(input_size, hidden_size).eval(), bidir_states, input_shapes_stacked),
+        ("lstm", lstm(input_size, hidden_size).eval(), states[0], input_shapes),
         (
-            stacked_bidir_lstm(input_size, hidden_size, num_layers).eval(),
-            stacked_bidir_states,
-            input_shapes_stacked_bidir,
+            "stacked",
+            stacked_lstm(input_size, hidden_size, num_layers).eval(),
+            states,
+            input_shapes_stacked,
         ),
+        ("bidir", bidir_lstm(input_size, hidden_size).eval(), bidir_states, input_shapes_stacked),
+        # TODO(masahi): stacked bidir seems to have a rare accuracy issue
+        # (
+        #     "stacked_bidir",
+        #     stacked_bidir_lstm(input_size, hidden_size, num_layers).eval(),
+        #     stacked_bidir_states,
+        #     input_shapes_stacked_bidir,
+        # ),
     ]
 
-    for (raw_model, states, input_shapes) in models:
+    for (name, raw_model, states, input_shapes) in models:
         script_module = torch.jit.script(raw_model)
         mod, params = from_pytorch(script_module, input_shapes)
 
@@ -356,4 +363,5 @@ def test_custom_lstm():
             params[states_name] = states_np
 
         for tgt, ctx in tvm.testing.enabled_targets():
+            print("Running %s on target %s" % (name, tgt))
             run_and_compare(mod, params, pt_result, target=tgt, ctx=ctx)