platform/aiml/mlops/docker/train/entrypoint.sh

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

echo "[train] Bootstrapping training container"
echo "[train] CUDA: $(python3 -c 'import torch;print(torch.version.cuda)') | Torch: $(python3 -c 'import torch;print(torch.__version__)')"

: "${MODEL_PATH:=/data/adaptai/platform/aiml/checkpoints/qwen3-8b-elizabeth-sft}"
: "${OUTPUT_DIR:=/data/ckpts/elizabeth}"
mkdir -p "$OUTPUT_DIR"

if [ ! -f "$MODEL_PATH/model.safetensors.index.json" ]; then
  echo "[train] MODEL_PATH missing. Set MODEL_PATH or mount /models."
fi

echo "[train] Ready. Override CMD to run your train script."
echo "Example: accelerate launch your_train.py --model $MODEL_PATH --output_dir $OUTPUT_DIR"
tail -f /dev/null