In class DatasetInputSpecsProvider when converting tf specs to jax
|
dtype=spec.dtype.as_numpy_dtype()) |
as_numpy_dtype is considered as a method when it is actually an attribute of tf.dtypes.Dtype (https://www.tensorflow.org/api_docs/python/tf/dtypes/DType#attributes).
The code works for most dtypes but fails to do for tf.string as the returned entity is a pointer to the object np datatype and not the object datatype itself.
In
class DatasetInputSpecsProviderwhen converting tf specs to jaxpraxis/praxis/base_input.py
Line 1116 in dd07dd6
as_numpy_dtypeis considered as a method when it is actually an attribute oftf.dtypes.Dtype(https://www.tensorflow.org/api_docs/python/tf/dtypes/DType#attributes).The code works for most dtypes but fails to do for
tf.stringas the returned entity is a pointer to the object np datatype and not theobjectdatatype itself.