Replace upstream library with ROCm/Wyoming deployment project

Remove original Kokoro library source, demo, examples, tests, JS port,
and GitHub config. Add Dockerfile (ROCm 6.1 / PyTorch 2.5.1), Wyoming
TCP server, docker-compose with GPU passthrough, config, entrypoint,
and Gitea Actions build workflow.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

entrypoint.sh

@@ -0,0 +1,26 @@
+#!/bin/bash
+set -e
+
+echo "=== Kokoro TTS Wyoming Server ==="
+
+# Show ROCm device info if available
+if command -v rocm-smi &>/dev/null; then
+    echo "--- ROCm Devices ---"
+    rocm-smi --showproductname 2>/dev/null || true
+    echo "--------------------"
+fi
+
+# Quick GPU availability check via Python
+python3 - <<'EOF'
+import torch
+available = torch.cuda.is_available()
+print(f"ROCm/CUDA available: {available}")
+if available:
+    count = torch.cuda.device_count()
+    for i in range(count):
+        print(f"  [{i}] {torch.cuda.get_device_name(i)}")
+else:
+    print("  WARNING: No GPU detected — running on CPU (performance will be degraded)")
+EOF
+
+exec python3 /app/server.py "$@"