TypeError: no default __reduce__ due to non-trivial __cinit__ #461
Labels
component:python sdk
Issue/PR related to Python SDK
status:triaged
Issue/PR triaged to the corresponding sub-team
type:bug
Something isn't working
Description of the bug:
Hi, I'm encourating this issue while using gemini-1.5-flash in DSPy. Not sure if it belongs here, but the issue only occurs with the Gemini model.
There's a line
program.deepcopy()
where a program with Gemini model is copied. (x). Program is a simple ChainOfThought with no demos.Actual vs expected behavior:
Other models are copied without an error at this stage.
Any other information you'd like to share?
google_generativeai-0.7.2
grpcio-status 1.48.1
Due to a protobuf error, tried both downgrading protobuf-4.25.3 to 3.19.6 as well as
os.environ["PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION"]="python"
and the issue persists.reductor
object that is giving an error is<built-in method __reduce_ex__ of grpc._cython.cygrpc.Channel object at 0x7f148c37e080>
The text was updated successfully, but these errors were encountered: