During my feature engingeering the following error occurred. My featurelist has 21 sublists with each 8537 values being either 0 or 1. When trying to run the One Hot Encoding via tensorflow it shows the error Could not find valid device for node
Does anyone have a quick fix for the error?
for feature in featurelist[1:]:
df = tensorflow.convert_to_tensor(feature, dtype=tensorflow.float32)
print(df)
df_enc = tensorflow.one_hot(df, 2, on_value=None, off_value=None, axis=None, dtype=None, name=None)
print(df_enc)
2020-05-29 15:08:41.969878: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x7fdefbc23d50 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-05-29 15:08:41.969919: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version
tf.Tensor([0. 0. 0. ... 0. 0. 0.], shape=(8537,), dtype=float32)
Traceback (most recent call last):
File "/Users/marius/Desktop/Masterarbeit/Github/virtual7/tempCodeRunnerFile.python", line 1234, in <module>
df_enc = tensorflow.one_hot(df, 2, on_value=None, off_value=None, axis=None, dtype=None, name=None)
File "/Users/marius/.pyenv/versions/3.7.3/lib/python3.7/site-packages/tensorflow_core/python/util/dispatch.py", line 180, in wrapper
return target(*args, **kwargs)
File "/Users/marius/.pyenv/versions/3.7.3/lib/python3.7/site-packages/tensorflow_core/python/ops/array_ops.py", line 3645, in one_hot
name)
File "/Users/marius/.pyenv/versions/3.7.3/lib/python3.7/site-packages/tensorflow_core/python/ops/gen_array_ops.py", line 5549, in one_hot
_ops.raise_from_not_ok_status(e, name)
File "/Users/marius/.pyenv/versions/3.7.3/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py", line 6606, in raise_from_not_ok_status
six.raise_from(core._status_to_exception(e.code, message), None)
File "<string>", line 3, in raise_from
tensorflow.python.framework.errors_impl.NotFoundError: Could not find valid device for node.
Node:{{node OneHot}}
All kernels registered for op OneHot :
device='XLA_CPU_JIT'; TI in [DT_INT32, DT_UINT8, DT_INT64]; T in [DT_FLOAT, DT_DOUBLE, DT_INT32, DT_UINT8, DT_INT16, ..., DT_UINT16, DT_COMPLEX128, DT_HALF, DT_UINT32, DT_UINT64]
device='XLA_CPU'; TI in [DT_INT32, DT_UINT8, DT_INT64]; T in [DT_FLOAT, DT_DOUBLE, DT_INT32, DT_UINT8, DT_INT16, ..., DT_UINT16, DT_COMPLEX128, DT_HALF, DT_UINT32, DT_UINT64]
device='CPU'; TI in [DT_UINT8]; T in [DT_INT64]
device='CPU'; TI in [DT_INT32]; T in [DT_INT64]
device='CPU'; TI in [DT_INT64]; T in [DT_INT64]
device='CPU'; TI in [DT_UINT8]; T in [DT_INT32]
.....
device='CPU'; TI in [DT_INT32]; T in [DT_INT32]
device='CPU'; TI in [DT_INT64]; T in [DT_INT32]
device='CPU'; TI in [DT_UINT8]; T in [DT_UINT16]
[Op:OneHot] name: one_hot/```
You casted your target to tf.float32
, which is incompatible with tf.one_hot
when using a list. You need to convert your target to an integer dtype before one-hot encoding. Try:
x = tf.cast(x, tf.int32)
Or, convert your tensor to a NumPy array:
tensor = np.array([0., 1., 2., 3.])
tf.one_hot(tensor, depth=4)
<tf.Tensor: shape=(4, 4), dtype=float32, numpy=
array([[1., 0., 0., 0.],
[0., 1., 0., 0.],
[0., 0., 1., 0.],
[0., 0., 0., 1.]], dtype=float32)>