Create start.sh

start.sh

@@ -0,0 +1,43 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+export HF_HUB_DISABLE_TELEMETRY=1
+MODEL_DIR="/app/bert_tf"
+mkdir -p "$MODEL_DIR"
+
+echo "[start] BioBERT bootstrap..."
+
+if [ -d "$MODEL_DIR" ] && [ -f "$MODEL_DIR/vocab.txt" ] && [ -f "$MODEL_DIR/config.json" ]; then
+  echo "[start] Model files already present in repo. Skipping download."
+else
+  if [ -n "${WEIGHTS_URL_TAR_GZ:-}" ]; then
+    # Single tar.gz containing the 5 files; supports Dropbox 'dl=1'
+    echo "[start] Downloading tarball from $WEIGHTS_URL_TAR_GZ"
+    curl -L "$WEIGHTS_URL_TAR_GZ" -o /tmp/bert_min.tar.gz
+    tar -xzf /tmp/bert_min.tar.gz -C "$MODEL_DIR"
+  elif [ -n "${WEIGHTS_BASE_URL:-}" ]; then
+    # Download files one by one from a base URL (must be direct links)
+    echo "[start] Downloading files from base $WEIGHTS_BASE_URL"
+    for f in checkpoint config.json vocab.txt model.ckpt-150000.index model.ckpt-150000.data-00000-of-00001 bert_config.json; do
+      echo "  - $f"
+      curl -fL "${WEIGHTS_BASE_URL%/}/$f" -o "$MODEL_DIR/$f" || true
+    done
+  else
+    echo "[start] ERROR: No weights source provided. Set WEIGHTS_URL_TAR_GZ or WEIGHTS_BASE_URL."
+    exit 1
+  fi
+fi
+
+# Ensure checkpoint file is correct
+if [ ! -f "$MODEL_DIR/checkpoint" ]; then
+  echo 'model_checkpoint_path: "model.ckpt-150000"' > "$MODEL_DIR/checkpoint"
+  echo 'all_model_checkpoint_paths: "model.ckpt-150000"' >> "$MODEL_DIR/checkpoint"
+fi
+
+# Fallback: if only bert_config.json exists, duplicate to config.json
+if [ ! -f "$MODEL_DIR/config.json" ] && [ -f "$MODEL_DIR/bert_config.json" ]; then
+  cp "$MODEL_DIR/bert_config.json" "$MODEL_DIR/config.json"
+fi
+
+echo "[start] Launching API on :${PORT:-7860}"
+exec uvicorn app:app --host 0.0.0.0 --port "${PORT:-7860}"