Hi, while using the following function,( KubeflowV2DagRunner ) I am facing an import error.tfx.v1.orchestration.experimental.KubeflowDagRunner | TFX | TensorFlow
Error- cannot import name KubeflowV2DagRunner from tfx.v1.orchestration.experimental
TFx version used- 1.13.0
The import error you’re experiencing is likely due to KubeflowV2DagRunner not being available in TFX version 1.13.0, possibly because it was introduced in a later version or has a different naming convention. You may need to update TFX or check the documentation for the correct class name.
Cheers,
Tim
KubeflowV2DagRunner is in orchestration/kubeflow/v2
isinstance(self.default_image, MutableMapping)
and self.default_image.get(_DEFAULT_IMAGE_PATH_KEY) is None
):
self.default_image[_DEFAULT_IMAGE_PATH_KEY] = _KUBEFLOW_TFX_IMAGE
if default_commands is None:
self.default_commands = KUBEFLOW_TFX_CMD
else:
self.default_commands = default_commands
class KubeflowV2DagRunner(tfx_runner.TfxRunner):
"""Kubeflow V2 pipeline runner (currently for managed pipelines).
Builds a pipeline job spec in json format based on TFX pipeline DSL object.
"""
def __init__(self,
config: KubeflowV2DagRunnerConfig,
output_dir: Optional[str] = None,
output_filename: Optional[str] = None):
"""Constructs an KubeflowV2DagRunner for compiling pipelines.