(base) [root@localhost ~]# docker exec finetune-trainer /opt/conda/bin/python -c "import torch; [print(f'GPU {i}: {torch.cuda.get_device_name(i)}, mem={torch.cuda.get_device_properties(i).total_memory/1e9:.2f}GB, alloc={torch.cuda.memory_allocated(i)/1e9:.2f}GB') for i in range(4)]"
GPU 0: MetaX N260, mem=68.48GB, alloc=0.00GB
GPU 1: MetaX N260, mem=68.48GB, alloc=0.00GB
GPU 2: MetaX N260, mem=68.48GB, alloc=0.00GB
GPU 3: MetaX N260, mem=68.48GB, alloc=0.00GB
(base) [root@localhost ~]# docker exec finetune-trainer /opt/conda/bin/python -c "import torch; print(torch.cuda.memory_allocated())"
0
(base) [root@localhost ~]# mx-smi 2>/dev/null || mcli-smi 2>/dev/null || echo "No smi tool found"
mx-smi  version: 2.2.9

=================== MetaX System Management Interface Log ===================
Timestamp                                         : Thu May 21 01:30:13 2026

Attached GPUs                                     : 4
+---------------------------------------------------------------------------------+
| MX-SMI 2.2.9                       Kernel Mode Driver Version: 3.4.4            |
| MACA Version: 3.3.0.15             BIOS Version: 1.30.0.0                       |
|------------------+-----------------+---------------------+----------------------|
| Board       Name | GPU   Persist-M | Bus-id              | GPU-Util      sGPU-M |
| Pwr:Usage/Cap    | Temp       Perf | Memory-Usage        | GPU-State            |
|==================+=================+=====================+======================|
| 0     MetaX N260 | 0           Off | 0000:b5:00.0        | 0%          Disabled |
| 53W / 225W       | 43C          P9 | 62108/65536 MiB     | Available            |
+------------------+-----------------+---------------------+----------------------+
| 1     MetaX N260 | 1           Off | 0000:b6:00.0        | 0%          Disabled |
| 49W / 225W       | 42C          P9 | 60952/65536 MiB     | Available            |
+------------------+-----------------+---------------------+----------------------+
| 2     MetaX N260 | 2           Off | 0000:b9:00.0        | 0%          Disabled |
| 53W / 225W       | 44C          P9 | 30691/65536 MiB     | Available            |
+------------------+-----------------+---------------------+----------------------+
| 3     MetaX N260 | 3           Off | 0000:bd:00.0        | 0%          Disabled |
| 51W / 225W       | 42C          P9 | 30469/65536 MiB     | Available            |
+------------------+-----------------+---------------------+----------------------+

+---------------------------------------------------------------------------------+
| Process:                                                                        |
|  GPU                    PID         Process Name                 GPU Memory     |
|                                                                  Usage(MiB)     |
|=================================================================================|
|  0                  1007916         VLLM::Worker_TP              59790          |
|  0                  1129825         python                       1618           |
|  1                  1007917         VLLM::Worker_TP              59790          |
|  1                  1129825         python                       490            |
|  2                   888820         VLLM::EngineCor              29530          |
|  2                  1129825         python                       490            |
|  3                   894310         VLLM::EngineCor              29180          |
|  3                  1129825         python                       618            |
+---------------------------------------------------------------------------------+

End of Log
(base) [root@loca