si264/era-directed-evolution
Official repository for datasets and experimental results for "Efficient, Few-shot Directed Evolution with Energy Rank Alignment".
2667
1#!/bin/bash2#SBATCH -N 13#SBATCH -G 44#SBATCH -C gpu&hbm80g5#SBATCH -q regular6#SBATCH -t 01:00:007#SBATCH -A m42358#SBATCH -J sampling9 10source /global/u2/s/si264/mambaforge/etc/profile.d/conda.sh11conda activate proera12 13target="TrpB4"14 15BASE_DIR="./$target/lightning_logs"16 17# Get a sorted list of subdirectories18subdirs=($(ls -d "$BASE_DIR"/*/ | sort))19 20# Run the script 10 times21for i in {0..9}; do22 # Extract the version number (subdirectory name)23 version_number=$(basename "${subdirs[$i]}")24 25 # Run the Python script with the specified arguments26 python sample_esm.py \27 --target $target \28 --num_samples 96 \29 --alignment_round 3 \30 --version_number "$version_number" \31 --replicate "$i"32done