feat: 推理与评测支持多卡 GPU 选择，训练任务实时 GPU 监控

- 新增 GPU 选择归一化 helper，统一前端各形态的选择（gpu_indices/gpus/gpu_id）
- 评测与推理支持同一节点内多卡选择，校验所选 GPU 空闲后再派发
- 新增 /fine-tune/{id}/gpu-status 接口，训练日志页展示实时 GPU 指标
- 算力节点推理加载支持 CUDA_VISIBLE_DEVICES 多卡可见，nvidia-smi 进程级监控
- GPU 占用跟踪细化为按卡记录，覆盖评测任务、推理模型与对比任务

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

backend/app/modules/compute_gateway/client.py

@@ -240,6 +240,10 @@ class ComputeNodeClient:
     async def inference_status(self) -> dict[str, Any]:
         return await self._request("GET", "/inference/status", timeout=INFERENCE_STATUS_TIMEOUT)
 
+    async def gpu_resources(self) -> list[dict[str, Any]]:
+        """Read live per-GPU metrics from this compute node."""
+        return await self.gpus()
+
     async def inference_unload(self) -> dict[str, Any]:
         return await self._request("POST", "/inference/unload", json_data={}, timeout=INFERENCE_UNLOAD_TIMEOUT)
 

backend/app/modules/compute_gateway/sync.py

@@ -73,7 +73,8 @@ async def reconcile_inference_loads(store: Any) -> list[dict[str, Any]]:
             if node_status == "ready":
                 item["status"] = "ready"
                 item.pop("error", None)
-                store.mark_inference_loaded(node["id"])
+                selected_gpus = item.get("gpu_indices") or item.get("gpus")
+                store.mark_inference_loaded(node["id"], selected_gpus)
             elif node_status == "error":
                 item["status"] = "error"
                 item["error"] = status.get("error") or "model load failed on compute node"