Digits and Weather Demos

README.md

@@ -1,3 +1,4 @@
+<<<<<<< HEAD
 # PARSL MPI
 
 This workflow runs OpenMPI hello world jobs using Parsl. The OpenMPI hello world code is defined in `mpitest.c` and, when compiled and executed, prints the hostname and rank in the format below:
@@ -84,3 +85,6 @@ No combination of `cores_per_worker`, `nodes_per_block` and `cores_per_node` ret
 2. Set `nodes_per_block` to the desired number of nodes per MPI task
 3. Set `cores_per_worker` to the number of cores in a single node
 4. Set `parallelism` to the desired number of nodes per MPI task
+=======
+This repository had been made for the sole purpose of testing DVC. I'm pretty sure I keep destroying my other one so imma start trying to be safer lol.
+>>>>>>> testing-repo/main

run_on_cluster/cvae-demos/digits-demo/.ipynb_checkpoints/README-checkpoint.md

@@ -0,0 +1,47 @@
+# CVAE Digits Example Demo
+This demo is dervied [Francois Chollet's](https://keras.io/examples/generative/vae/) Variational AutoEncoder example, modified for online learning. In particular, some goals are:
+
+* Test computational performance scaling (CPUs, GPUs)
+* Test model performance while saving and fitting again (in particular, adding more data instead of having all the data available at once).
+* Integrations with DVC
+
+## Environment Set Up
+To get started with the CVAE Digit Model, follow the instructions below to set up your environment.
+
+### Create Environment:
+```
+source ~/pw/.miniconda3c/etc/profile.d/conda.sh
+conda create --name <NAME> python=3.9
+conda activate <NAME>
+```
+
+### Install Packages:
+The following packages need to be installed on top of a typical base Conda env. Install the packages in the following order so the environment solves correctly:
+```
+conda install -y -c conda-forge tensorflow
+conda install -y -c conda-forge matplotlib
+conda install -y -c conda-forge pandas
+conda install -y -c conda-forge dvc 
+
+pip install mlflow
+```
+
+### Connect Notebook to Environmant:
+```
+conda install -y ipykernel
+conda install -y requests
+conda install -y -c anaconda jinja2
+python -m ipykernel install --user --name=<NAME> --display-name "Python (<NAME>)"
+```
+
+### Alternative Set Up:
+
+1. Run `conda env update --file digits_env.yaml --name <NAME>` to install the conda environment with all the needed dependencies 
+2. Run `source ~/pw/.miniconda3c/etc/profile.d/conda.sh` to initiate conda 
+3. Run `conda activate <NAME>` to activate the environment you installed
+4. Run `python -m ipykernel install --user --name=<NAME> --display-name "Python (<NAME>)"` to connect the environment to the notebook
+
+## Needed Specifications
+
+**Storage:** Mounted stoarge is required for dvc to run properly.\
+**GitHub:** A separate git repo that can be SSH into is also needed for dvc to run properly. SSH connection must be established on the terminal.