jelc53

# ensure conda-forge is set as the priority channel
conda config --add channels conda-forge
conda config --set channel_priority strict

# create conda env 
conda create -c conda-forge -c nodefaults -n bayes_env "pymc=5.10" "numpy=1.25.2" "scipy=1.12.0" python=3.10.6
conda activate bayes_env
pip install numba
conda install -c conda-forge mamba
mamba install -c conda-forge nutpie

jelc53

# Make sure latex installed (example for mac below)
brew install --cask mactex

# Install texlive package to interpret sphinx rst docs
brew install texlive

# Open TeX Live Utility and install for "latexmk"
Tex Live Utility > latexmk > install

# (Option 1) You should now be able to run the following command from the sphinx docs directory

jelc53

Preparation: remove dual-boot, download theme images and conf repo to USB

• Removing dual-boot: removing Windows from my lenovo dual-boot machine from both disk and grub menu https://forums.linuxmint.com/viewtopic.php?t=386957. To do this, install gparted (sudo apt install gparted) and unmount, delete and resize partitions from OS partition you want to hold onto (e.g. Ubuntu)
• In the early stages of installation, access to github is limited and so best to store conf and any theme images on a USB you can mount and access from root terminal straight away

Installation: boot from live iso image of void (flash USB) and run void-installer wizard

• Download void image: sha256sum.sig, sha256sum.txt and live version of void that matches operating system (base and non-musl) https://repo-default.voidlinux.org/live/current/
• Prepare USB device: plug in USB, find sdX reference with fdisk -l, unmount with umount /dev/sdX, and write live image with `dd bs=4M if=/path/to/void-live-ARCH-DATE-VARIANT.iso

jelc53

Finetuning processing

Step 1 (Optional): Train Classification Model
- data_dir = '/home/ubuntu/deepsolar/data/bdappv-france/ft_100/'
- old_ckpt_path = '/home/ubuntu/deepsolar/models/deepsolar_pretrained.pth'
- ckpt_save_dir = 'checkpoint/bdappv_ft100_w0.01'
- model_name = 'deepsolar_classification'
- imbalance_rate = 1
- learning_rate = 0.01
- weight_decay = 0.01

jelc53

#!/usr/bin/env bash

## The destination folder where your files will
## be copied to.
dest="bar";

## For each file path in your input file
while read path; do 
    ## $target is the name of the file, removing the path. 
    ## For example, given /foo/bar.txt, the $target will be bar.txt.

jelc53

import json
import os
from glob import glob
from collections import defaultdict
from argparse import ArgumentParser
import pandas as pd

TIME_LIMIT = 600

jelc53

# download french dataset and unzip in data directory
# sub dir structure: data/bdappv-france/bdappv/[ign,google]/[img,mask]
mkdir bdappv-france
unzip bdappv.zip

# checkout retrain_pytorch github branch
git checkout retrain_pytorch
git pull

# create "dummy" data sub-directory for testing

jelc53

# checkout main git branch
git pull
git checkout master

# download and scp coqgym's astactic best model into model directory
# https://drive.google.com/drive/folders/1AzLaEpoGS3BPMUz9Bl63MHAFRqlF4CtH
scp model.pth [username]@scq2.stanford.edu:/scr2/[user_dir]/CoqGym-GNN/ASTactic/runs/astactic/checkpoints/model.pth

# evaluation runsheet (5 left to go!)
    "weak-up-to",                 -- done

jelc53

#!/bin/bash

sudo apt-get install git python3-pip python-is-python3 make build-essential libssl-dev zlib1g-dev libbz2-dev libreadline-dev libsqlite3-dev curl libffi-dev
pip install virtualenvwrapper

git clone https://github.com/pyenv/pyenv.git ~/.pyenv
git clone https://github.com/pyenv/pyenv-virtualenv.git ~/.pyenv/plugins/pyenv-virtualenv
git clone https://github.com/pyenv/pyenv-virtualenvwrapper.git ~/.pyenv/plugins/pyenv-virtualenvwrapper

cat << EOF >> ~/.bashrc

jelc53

def max_pool_forward_naive(x, pool_param):
  """A naive implementation of the forward pass for a max-pooling layer.

  Inputs:
  - x: Input data, of shape (N, C, H, W)
  - pool_param: dictionary with the following keys:
    - 'pool_height': The height of each pooling region
    - 'pool_width': The width of each pooling region
    - 'stride': The distance between adjacent pooling regions