Manufacturer:HuaweiProduct Name: A800T A2
Ascend-cann-toolkit: 8.0.RC3.alpha001_linux-aarch64
npu-smi version: 23.0.3

Train:   0%|                                                                                                                         | 0/3095 [00:00<?, ?it/s][rank0]:[E compiler_depend.ts:270] call aclnnArgMax failed, detail:E89999: Inner Error!
E89999: 2024-10-08-10:59:27.761.810  op[ArgMaxV2], TilingPrepareForNLLLoss, get is_vsel_support error[FUNC:TilingPrepareCommon][FILE:arg_max_v2.cc][LINE:513]
        TraceBack (most recent call last):
        op[ArgMaxV2], TilingPrepareCommon is failed[FUNC:TilingPrepareForArgOps][FILE:arg_max_v2.cc][LINE:532]
        Op tiling compile info parse failed
        InitTilingParseCtx failed
        Kernel Run failed. opType: 37, ArgMaxV2
        launch failed for ArgMaxV2, errno:361001.

[ERROR] 2024-10-08-10:59:27 (PID:158327, Device:0, RankID:0) ERR01005 OPS internal error
Exception raised from operator() at build/CMakeFiles/torch_npu.dir/compiler_depend.ts:41 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x68 (0xffff8213aee8 in /root/anaconda3/envs/swift-npu/lib/python3.10/site-packages/torch/lib/libc10.so)