#!/bin/bash
#SBATCH --job-name=dovla_12task_train
#SBATCH --nodes=1
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=8
#SBATCH --gres=gpu:1
#SBATCH --mem=64000M
#SBATCH --time=96:00:00
#SBATCH --output=logs/phase_b_12task_train_%A_%a.out
#SBATCH --error=logs/phase_b_12task_train_%A_%a.err
#SBATCH --array=0-2

set -euo pipefail

# Phase B: Train on 12-task collection (3 seeds)

PROJECT_DIR="${PROJECT_DIR:-$PWD}"
cd "$PROJECT_DIR"

source .venv/bin/activate

DATASET="/scratch/$USER/dovla/experiments/phase_b_12task_collection/merged_12tasks"
OUT_DIR="/scratch/$USER/dovla/experiments/phase_b_12task_model"
SEED=$SLURM_ARRAY_TASK_ID

mkdir -p "$OUT_DIR/seed_$SEED" logs

echo "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "="
echo "Phase B: Training on 12-Task Collection"
echo "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "=" "="
echo ""
echo "Seed: $SEED"
echo "Tasks: 12 (6 existing + 6 new)"
echo "Groups: ~6,200"
echo "Hidden dim: 1024 (larger for 12 tasks)"
echo ""

python scripts/train_dovla.py \
    --dataset "$DATASET" \
    --out "$OUT_DIR/seed_$SEED" \
    --objective lattice_field \
    --hidden-dim 1024 \
    --action-horizon 4 \
    --epochs 100 \
    --batch-groups 16 \
    --records-per-group 8 \
    --lr 0.0003 \
    --weight-decay 0.01 \
    --dropout 0.1 \
    --warmup-steps 1000 \
    --device auto \
    --seed $SEED \
    --observation-mode state \
    --loss-weight bc=1.0 \
    --loss-weight field_effect=1.0 \
    --loss-weight field_utility_regression=1.0 \
    --loss-weight field_utility_margin=0.5 \
    --loss-weight field_preference=0.5 \
    --loss-weight effect_anchor=0.1

echo ""
echo "✅ Phase B training complete (seed $SEED)"
echo ""
echo "Next: Evaluate on held-out tasks"