Closed umnooob closed 4 months ago
Thanks @umnooob for raising this issue.
Looking at your backtrace I see the following code fragment from vector_service_pb2.py
[/usr/local/lib/python3.10/dist-packages/pinecone/core/grpc/protos/vector_service_pb2.py](https://localhost:8080/#) in <module>
39 create_key=_descriptor._internal_create_key,
40 fields=[
---> 41 _descriptor.FieldDescriptor(
42 name='indices', full_name='SparseValues.indices', index=0,
43 number=1, type=13, cpp_type=3, label=3,
However, we can see that lines 39-43 do not look like this in v4.0.0 of pinecone-client
: https://github.com/pinecone-io/pinecone-python-client/blob/06c69fbbe5c3fa57717ba71596d94f03ee50aaa3/pinecone/core/grpc/protos/vector_service_pb2.py#L39-L43
Indeed, the lines in your backtrace match the v3.2.2 code: https://github.com/pinecone-io/pinecone-python-client/blob/c5ba9ce0adb31cde7a7b779d0bdfaaf186f596ec/pinecone/core/grpc/protos/vector_service_pb2.py#L39-L43
I believe there is an issue with your Python environment and v4.0.0 has not been installed correctly - I suggest you start with a clean environment (for example using venv) and reinstall your packages.
Indeed, if I attempt to reproduce the issue you see I can't even get pip to install pinecone-client
v4.0.0 and the necessary LlamaIndex packages as they do not support v4.0.0 yet:
❯ python3.10 -m venv .venv
❯ source .venv/bin/activate
❯ pip install "pinecone-client[grpc]==4.0.0" llama-index-vector-stores-pinecone
Collecting pinecone-client==4.0.0 (from pinecone-client[grpc]==4.0.0)
Using cached pinecone_client-4.0.0-py3-none-any.whl.metadata (16 kB)
Collecting llama-index-vector-stores-pinecone
Using cached llama_index_vector_stores_pinecone-0.1.6-py3-none-any.whl.metadata (674 bytes)
...
INFO: pip is looking at multiple versions of llama-index-vector-stores-pinecone to determine which version is compatible with other requirements. This could take a while.
Collecting llama-index-vector-stores-pinecone
...
ERROR: Cannot install llama-index-vector-stores-pinecone==0.0.1, llama-index-vector-stores-pinecone==0.1.0, llama-index-vector-stores-pinecone==0.1.1, llama-index-vector-stores-pinecone==0.1.2, llama-index-vector-stores-pinecone==0.1.3, llama-index-vector-stores-pinecone==0.1.4, llama-index-vector-stores-pinecone==0.1.5, llama-index-vector-stores-pinecone==0.1.6, pinecone-client==4.0.0 and pinecone-client[grpc]==4.0.0 because these package versions have conflicting dependencies.
The conflict is caused by:
The user requested pinecone-client==4.0.0
pinecone-client[grpc] 4.0.0 depends on pinecone-client 4.0.0 (from https://files.pythonhosted.org/packages/9c/64/081b55a33e492fc181524a955c2b65ba8a628dbc1bb897e65b723c7b7ffc/pinecone_client-4.0.0-py3-none-any.whl (from https://pypi.org/simple/pinecone-client/) (requires-python:<4.0,>=3.8))
llama-index-vector-stores-pinecone 0.1.6 depends on pinecone-client<4.0.0 and >=3.0.2
Is this a new bug in the Pinecone Python client?
Current Behavior
When i use the newest Pinecone Python client 4.0, it comes to me:
Expected Behavior
when i use Pinecone Python client == 3.2.2, everything ok for me. I believe this pr https://github.com/pinecone-io/pinecone-python-client/pull/334 should have introduced bugs
Steps To Reproduce
pip install pinecone-client[grpc]==4.0.0
Relevant log output
No response
Environment
Additional Context
No response