Add runtime entrypoint for Gemma 4 API

entrypoint.sh

@@ -0,0 +1,19 @@
+#!/bin/sh
+set -eu
+
+MODEL_SPEC="${MODEL_SPEC:-unsloth/gemma-4-E2B-it-GGUF:Q4_0}"
+HOST="${HOST:-0.0.0.0}"
+PORT="${PORT:-7860}"
+CTX_SIZE="${CTX_SIZE:-4096}"
+THREADS="${THREADS:-2}"
+PARALLEL="${PARALLEL:-1}"
+REASONING_MODE="${REASONING_MODE:-off}"
+
+exec llama-server \
+  -hf "$MODEL_SPEC" \
+  --host "$HOST" \
+  --port "$PORT" \
+  --ctx-size "$CTX_SIZE" \
+  --threads "$THREADS" \
+  --parallel "$PARALLEL" \
+  --reasoning "$REASONING_MODE"