You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by GitBox <gi...@apache.org> on 2020/04/21 06:07:44 UTC
[GitHub] [incubator-tvm] u99127 opened a new pull request #5392: [WIP] Migrate tf to 1.15 and 2.1
u99127 opened a new pull request #5392:
URL: https://github.com/apache/incubator-tvm/pull/5392
This shows the changes needed to CI and the testsuite to get CI to support TF/TFLite 1.15.2 and then on to TF / TFlite 2.1.0
There were a couple of related issues found with Tensorflow 2.1.0 which are fixed with #5391 . However issues still remain.
The 3 test failures in TFlite testsuite need to be investigated. Either they have regresssed since support was added or they have always been the case. I'll look to investigate them and create issues for them in the issue tracker.
There are still 2 open issues with running Tensorflow tests with 2.1.0 which need investigation.
FAILED tests/python/frontend/tensorflow/test_forward.py::test_tensor_array_size
FAILED tests/python/frontend/tensorflow/test_forward.py::test_tensor_array_unstack
tests/python/frontend/tensorflow/test_forward.py:950: in run
compare_tf_with_tvm([], [], 'TensorArraySizeV3:0', mode='debug')
tests/python/frontend/tensorflow/test_forward.py:178: in compare_tf_with_tvm
final_graph_def = tf_testing.AddShapesToGraphDef(sess, out_node)
python/tvm/relay/testing/tf.py:95: in AddShapesToGraphDef
convert_to_list(out_node),
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/util/deprecation.py:324: in new_func
return func(*args, **kwargs)
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/graph_util_impl.py:277: in convert_variables_to_constants
inference_graph = extract_sub_graph(input_graph_def, output_node_names)
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/util/deprecation.py:324: in new_func
return func(*args, **kwargs)
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/graph_util_impl.py:197: in extract_sub_graph
_assert_nodes_are_present(name_to_node, dest_nodes)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
name_to_node = {'Const': name: "Const"
op: "Const"
attr {
key: "_output_shapes"
value {
list {
shape {
}
}
... key: "value"
value {
tensor {
dtype: DT_INT32
tensor_shape {
}
int_val: 2
}
}
}
}
nodes = ['TensorArraySizeV3']
def _assert_nodes_are_present(name_to_node, nodes):
"""Assert that nodes are present in the graph."""
for d in nodes:
> assert d in name_to_node, "%s is not in graph" % d
E AssertionError: TensorArraySizeV3 is not in graph
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/graph_util_impl.py:152: AssertionError
__________________________ test_tensor_array_unstack ___________________________
def test_tensor_array_unstack():
def run(dtype_str, input_shape, infer_shape):
with tf.Graph().as_default():
dtype = tf_dtypes[dtype_str]
t = tf.constant(np.random.choice([0, 1, 2, 3],
size=input_shape).astype(dtype.name))
ta1 = tf.TensorArray(dtype=dtype, infer_shape=infer_shape, size=input_shape[0])
ta2 = ta1.unstack(t)
out0 = ta2.size()
out1 = ta2.read(0)
compare_tf_with_tvm([], [], 'TensorArraySizeV3:0', mode='debug')
compare_tf_with_tvm([], [], 'TensorArrayReadV3:0', mode='debug')
for dtype in ["float32", "int8"]:
> run(dtype, (5,), False)
tests/python/frontend/tensorflow/test_forward.py:986:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
tests/python/frontend/tensorflow/test_forward.py:983: in run
compare_tf_with_tvm([], [], 'TensorArraySizeV3:0', mode='debug')
tests/python/frontend/tensorflow/test_forward.py:178: in compare_tf_with_tvm
final_graph_def = tf_testing.AddShapesToGraphDef(sess, out_node)
python/tvm/relay/testing/tf.py:95: in AddShapesToGraphDef
convert_to_list(out_node),
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/util/deprecation.py:324: in new_func
return func(*args, **kwargs)
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/graph_util_impl.py:277: in convert_variables_to_constants
inference_graph = extract_sub_graph(input_graph_def, output_node_names)
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/util/deprecation.py:324: in new_func
return func(*args, **kwargs)
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/graph_util_impl.py:197: in extract_sub_graph
_assert_nodes_are_present(name_to_node, dest_nodes)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
name_to_node = {'Const': name: "Const"
op: "Const"
attr {
key: "_output_shapes"
value {
list {
shape {
dim {
...: "value"
value {
tensor { dtype: DT_INT32
tensor_shape {
}
int_val: 5
}
}
}
, ...}
nodes = ['TensorArraySizeV3']
def _assert_nodes_are_present(name_to_node, nodes):
"""Assert that nodes are present in the graph."""
for d in nodes:
> assert d in name_to_node, "%s is not in graph" % d
E AssertionError: TensorArraySizeV3 is not in graph
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-tvm] u99127 commented on pull request #5392: [CI] Migrate Tensorflow and Tensorflow lite in CI to 2.1.0
Posted by GitBox <gi...@apache.org>.
u99127 commented on pull request #5392:
URL: https://github.com/apache/incubator-tvm/pull/5392#issuecomment-618582877
Gentle ping .
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-tvm] tqchen commented on pull request #5392: [CI] Migrate Tensorflow and Tensorflow lite in CI to 2.1.0
Posted by GitBox <gi...@apache.org>.
tqchen commented on pull request #5392:
URL: https://github.com/apache/incubator-tvm/pull/5392#issuecomment-621363701
NOTE: the mainline ci-gpu now takes in effect
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-tvm] tqchen commented on pull request #5392: [CI] Migrate Tensorflow and Tensorflow lite in CI to 2.1.0
Posted by GitBox <gi...@apache.org>.
tqchen commented on pull request #5392:
URL: https://github.com/apache/incubator-tvm/pull/5392#issuecomment-618604193
This PR is merged. WIll update the thread when the CI binary get updated
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org