New RunLLM response generated: >>>*# Question* (a...
# runllm-flyte-trial
j
New RunLLM response generated: >>>*# Question* (absolute rookie writing) I tried to run the code from the quickstart guide and it works fine on my machine. Now I would like to run it in a dev container and some how it fails. No clue why. Can you help me? The error it throws is: (flyte-example) root@5bd93e9b530a:~/miniconda3/bin# pyflyte run example.py hello_world_wf ╭────────────────────────────────────────────────────── Traceback (most recent call last) ──────────────────────────────────────────────────────╮ │ /root/miniconda3/envs/flyte-example/bin/pyflyte:5 in <module> │ │ │ │ ❱ 5 from flytekit.clis.sdk_in_container.pyflyte import main │ │ │ │ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/flytekit/clis/sdk_in_container/pyflyte.py:7 in <module> │ │ │ │ ❱ 7 from flytekit.clis.sdk_in_container.backfill import backfill │ │ │ │ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/flytekit/clis/sdk_in_container/backfill.py:6 in <module> │ │ │ │ ❱ 6 from flytekit.clis.sdk_in_container.helpers import get_and_save_remote_with_click_contex │ │ │ │ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/flytekit/clis/sdk_in_container/helpers.py:8 in <module> │ │ │ │ ❱ 8 from flytekit.configuration.plugin import get_plugin │ │ │
>>│ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/flytekit/configuration/plugin.py:28 in <module> │
│ │ │ ❱ 28 from flytekit.remote import FlyteRemote │ │ │ │ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/flytekit/remote/__init__.py:99 in <module> │ │ │ │ ❱ 99 from flytekit.remote.remote import FlyteRemote │ │ │ │ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/flytekit/remote/remote.py:35 in <module> │ │ │ │ ❱ 35 from flytekit.clients.friendly import SynchronousFlyteClient │ │ │ │ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/flytekit/clients/friendly.py:16 in <module> │ │ │ │ ❱ 16 from flyteidl.service import dataproxy_pb2 as _data_proxy_pb2 │ │ │ │ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/flyteidl/service/dataproxy_pb2.py:15 in <module> │ │ │ │ ❱ 15 from protoc_gen_openapiv2.options import annotations_pb2 as protoc__gen__openapiv2_dot_o │ │ │
>>│ /root/miniconda3/envs/flyte-example/lib/python3.10/site-packages/protoc_gen_openapiv2/options/annotations_pb2.py:23 in <module> │
│ │ │ ❱ 23 google_dot_protobuf_dot_descriptor__pb2.FileOptions.RegisterExtension(openapiv2_swagge │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ AttributeError: type object 'FileOptions' has no attribute 'RegisterExtension' The requirements.txt for the dev container is: torch pytorch-lightning torchvision lightning fvcore iopath notebook ipywidgets scikit-learn flytekit flytekitplugins-kfpytorch flytekitplugins-deck-standard dvc And the base image is: nvcr.io/nvidia/pytorch:23.12-py3 With the following modifications: COPY requirements.txt /tmp/requirements.txt RUN pip install -r /tmp/requirements.txt RUN pip install "git+https://github.com/facebookresearch/pytorch3d.git@stable" RUN curl -sL https://ctl.flyte.org/install | bash -s -- -b /usr/local/bin COPY .flyte/config.yaml /root/.flyte/config.yaml RUN rm /tmp/requirements.txt # Answer The error you're encountering,
AttributeError: type object 'FileOptions' has no attribute 'RegisterExtension'
, suggests a compatibility issue with the protobuf library. This is a common issue when the installed version of protobuf is not compatible with the other packages in your environment, particularly those related to Flyte. 1. Check Protobuf Version: Ensure that the protobuf version is compatible with Flyte and other installed packages. Flyte often requires a specific range of protobuf versions. You can specify a compatible version in your
requirements.txt
file, for example,
protobuf==3.20.3
. 2. Rebuild the Container: After updating the
requirements.txt
with the correct protobuf version, rebuild your Docker container to ensure all dependencies are correctly installed. 3. Verify Dependency Conflicts: Check for any other dependency conflicts that might arise from the packages listed in your
requirements.txt
. You can use tools like
pipdeptree
to visualize and resolve any conflicts. Would you like more detailed steps on how to resolve this issue or further assistance with Flyte setup? Sources: - Flyte GitHub Issue #3006 Slack Link https://.slack.com/archives/CP2HDHKE1/p1729286082.525619 👍 0 button 👎 0 button 💡 Hint: Mention @runllm in the thread for followups.