scripts/eval_cavi_epoch20_gpu0.sh

#!/usr/bin/env bash
set -euo pipefail

PROJECT_DIR=/data/DAIDASEN/ViSA/CAVI-ReID
PYTHON_BIN=/data/conda/cavi_clip/bin/python
CHECKPOINT="$PROJECT_DIR/logs/CARGO/CLIP_CAVI_JOINT_60E_B64_V2/model_epoch_20.pth"
OUTPUT_DIR="$PROJECT_DIR/logs/CARGO/CLIP_CAVI_EPOCH20_EVAL"
LAUNCH_LOG="$PROJECT_DIR/logs/CARGO/clip_cavi_epoch20_eval_launcher.log"

mkdir -p "$OUTPUT_DIR"
echo "$(date -Is) waiting for epoch 20 checkpoint and GPU 0" >> "$LAUNCH_LOG"
while [[ ! -s "$CHECKPOINT" ]]; do sleep 30; done
sleep 30
while true; do
  used_mib=$(nvidia-smi --query-gpu=memory.used --format=csv,noheader,nounits -i 0 | tr -d ' ')
  if [[ "$used_mib" =~ ^[0-9]+$ ]] && (( used_mib < 2000 )); then break; fi
  sleep 30
done

echo "$(date -Is) starting epoch 20 evaluation on GPU 0" >> "$LAUNCH_LOG"
cd "$PROJECT_DIR"
export CUDA_VISIBLE_DEVICES=0
export FASTREID_DATASETS=/data/DAIDASEN/datasets
exec "$PYTHON_BIN" baselines/hihr/train.py \
  --eval-only \
  --checkpoint "$CHECKPOINT" \
  --config-file configs/CARGO/clip_cavi.yml \
  --output-dir "$OUTPUT_DIR" \
  --device cuda:0 \
  >> "$OUTPUT_DIR/eval.log" 2>&1