SY2567
January 19, 2023, 9:58pm
1
When ray is being initialized through
if not ray.is_initialized():
ray.init()
or
tuner.fit()
The following is printed and it does not complete the rest of the script,
INFO:ray.tune.tune:Initializing Ray automatically.For cluster usage or custom Ray initialization, call ray.init(...)
before tune.run
.
2023-01-19 07:04:42,868 INFO worker.py:1518 – Started a local Ray instance.
SY2567
January 31, 2023, 7:26am
2
I am using a dockerfile:
FROM python:3.7-slim
RUN pip3 install tensorflow==2.6.2 tensorflow-io==0.27.0
FROM jupyter/base-notebook:python-3.7.6
ENV TRANSFORMERS_CACHE=/tmp/.cache
ENV TOKENIZERS_PARALLELISM=true
USER root
RUN apt-get update && apt-get install -y libglib2.0-0
RUN pip3 install --no-cache-dir protobuf==3.20.1
RUN pip3 install --no-cache-dir gym==0.23.1
RUN pip3 install --no-cache-dir scikit-learn
RUN pip3 install --no-cache-dir importlib-metadata==4.12.0
RUN pip3 install --no-cache-dir ray[rllib]==2.0.0 torch==1.9.0
RUN pip3 install --no-cache-dir importlib-resources ale-py~=0.7.1 \
&& pip3 install --no-cache-dir MarkupSafe==2.0.1
USER ${NB_UID}
WORKDIR "${HOME}/work"