在提交新问题之前，请确保您已经在社区中搜索过相关问题，并使用了社区中提供的资源/工具后，仍未找到满意的解决方式。

⚠️ 安全信息提醒：请仔细检查提供的文本内容，确保其不包含敏感数据信息，包括但不限于：

• API 令牌或密钥
• 密码或身份验证凭证
• 私有网址或接口地址
• 个人或机密数据
• ...

在分享配置信息或代码示例时，请将敏感信息脱敏处理，或使用 <TOKEN> 等占位符替代原有内容。

环境信息

例如：
- 操作系统
- 昇腾硬件信息
- CANN软件版本
- 安装的对应软件版本

🐛 问题描述

npu  train soft_actor_critic                  
learning rate: 0.01
[W716 16:38:17.155472250 ToKernelNpu.cpp:40] Warning: Device do not support double dtype now, dtype cast replace with float. (function operator())
tolerance: 0.001
[eager] step: 1 step_time: 1.7530699999497301 ms loss: -3.712211847305298
[eager] step: 2 step_time: 1.4726500000961096 ms loss: -256.0
[eager] step: 3 step_time: 1.4775000000781802 ms loss: -256.0
[eager] step: 4 step_time: 1.5320599999313345 ms loss: -256.0
[eager] step: 5 step_time: 1.4531299999589464 ms loss: -256.0
[eager] step: 6 step_time: 1.4148199999226563 ms loss: -256.0
[eager] step: 7 step_time: 1.4243300000771342 ms loss: -256.0
[eager] step: 8 step_time: 1.5997900000002119 ms loss: -256.0
[eager] step: 9 step_time: 1.4486699999451957 ms loss: -256.0
[eager] step: 10 step_time: 1.3574600000083592 ms loss: -256.0
[eager] step: 11 step_time: 1.4221799999631912 ms loss: -256.0
[eager] step: 12 step_time: 1.3556000000107815 ms loss: -256.0
[eager] step: 13 step_time: 1.3707299999623501 ms loss: -256.0
[eager] step: 14 step_time: 1.4106099999935395 ms loss: -256.0
[eager] step: 15 step_time: 1.4358700000229874 ms loss: -256.0
[eager] step: 16 step_time: 1.4673100000663908 ms loss: -256.0
[eager] step: 17 step_time: 1.405249999947955 ms loss: -256.0
[eager] step: 18 step_time: 1.3857300000381656 ms loss: -256.0
[eager] step: 19 step_time: 1.4345700000149009 ms loss: -256.0
[eager] step: 20 step_time: 1.4077399999905538 ms loss: -256.0
[eager] step: 21 step_time: 1.4034199999741759 ms loss: -256.0
[eager] step: 22 step_time: 1.7552499999737847 ms loss: -256.0
[eager] step: 23 step_time: 1.4311500000303568 ms loss: -256.0
[eager] step: 24 step_time: 1.4430799999445298 ms loss: -256.0
[eager] step: 25 step_time: 1.392939999959708 ms loss: -256.0
[eager] step: 26 step_time: 1.3780499999711537 ms loss: -256.0
[eager] step: 27 step_time: 1.3497899999492802 ms loss: -256.0
[eager] step: 28 step_time: 1.3604800000166506 ms loss: -256.0
[eager] step: 29 step_time: 1.4156199999888486 ms loss: -256.0
[eager] step: 30 step_time: 1.353989999984151 ms loss: -256.0
[eager] step: 31 step_time: 1.3741799999706927 ms loss: -256.0
[eager] step: 32 step_time: 1.3739600000235441 ms loss: -256.0
[eager] step: 33 step_time: 1.3824999999769716 ms loss: -256.0
[eager] step: 34 step_time: 1.3947600000392413 ms loss: -256.0
[eager] step: 35 step_time: 1.4127900000175941 ms loss: -256.0
[eager] step: 36 step_time: 1.4000300000134303 ms loss: -256.0
[eager] step: 37 step_time: 1.4442399999552435 ms loss: -256.0
[eager] step: 38 step_time: 1.3579499999423206 ms loss: -256.0
[eager] step: 39 step_time: 1.4230000000452492 ms loss: -256.0
[eager] step: 40 step_time: 1.3158999998950094 ms loss: -256.0
[eager] step: 41 step_time: 1.3232500000412983 ms loss: -256.0
[eager] step: 42 step_time: 1.3441400000147041 ms loss: -256.0
[eager] step: 43 step_time: 1.353730000005271 ms loss: -256.0
[eager] step: 44 step_time: 1.3458199999831777 ms loss: -256.0
[eager] step: 45 step_time: 1.3476799999807554 ms loss: -256.0
[eager] step: 46 step_time: 1.2886700000080964 ms loss: -256.0
[eager] step: 47 step_time: 1.3534500000105254 ms loss: -256.0
[eager] step: 48 step_time: 1.3255799999569717 ms loss: -256.0
[eager] step: 49 step_time: 1.3102200000503217 ms loss: -256.0
[eager] step: 50 step_time: 1.3920700000653596 ms
[eager] summary [16-40] total steps time: 35.26 ms, avg step time: 1.41 ms
------------------------------------------------------------------------------------------------------
[compile] step: 1 step_time: 19680.90195000002 ms loss: -3.712211847305298
[compile] step: 2 step_time: 242.09932999997363 ms loss: -256.0
[compile] step: 3 step_time: 3.166120000059891 ms loss: -256.0
[compile] step: 4 step_time: 2.450049999993098 ms loss: -256.0
[compile] step: 5 step_time: 1.808080000046175 ms loss: -256.0
[compile] step: 6 step_time: 1.7572199999449367 ms loss: -256.0
[compile] step: 7 step_time: 1.6714900000351918 ms loss: -256.0
[compile] step: 8 step_time: 1.7485100000840248 ms loss: -256.0
[compile] step: 9 step_time: 1.7294900000024427 ms loss: -256.0
[compile] step: 10 step_time: 1.6257199999927252 ms loss: -256.0
[compile] step: 11 step_time: 1.6323500000225977 ms loss: -256.0
[compile] step: 12 step_time: 1.6464999999925567 ms loss: -256.0
[compile] step: 13 step_time: 1.626170000008642 ms loss: -256.0
[compile] step: 14 step_time: 1.646479999976691 ms loss: -256.0
[compile] step: 15 step_time: 1.605240000003505 ms loss: -256.0
[compile] step: 16 step_time: 1.699209999969753 ms loss: -256.0
[compile] step: 17 step_time: 1.6288100000565464 ms loss: -256.0
[compile] step: 18 step_time: 1.6089499999907275 ms loss: -256.0
[compile] step: 19 step_time: 1.6337800000201241 ms loss: -256.0
[compile] step: 20 step_time: 1.6594600000416904 ms loss: -256.0
[compile] step: 21 step_time: 1.5732000000525659 ms loss: -256.0
[compile] step: 22 step_time: 1.5216099999406651 ms loss: -256.0
[compile] step: 23 step_time: 1.5297600000394596 ms loss: -256.0
[compile] step: 24 step_time: 1.5533199999708813 ms loss: -256.0
[compile] step: 25 step_time: 1.6374900000073467 ms loss: -256.0
[compile] step: 26 step_time: 1.674839999964206 ms loss: -256.0
[compile] step: 27 step_time: 1.6193600000633523 ms loss: -256.0
[compile] step: 28 step_time: 1.620590000015909 ms loss: -256.0
[compile] step: 29 step_time: 1.6226700000743222 ms loss: -256.0
[compile] step: 30 step_time: 1.616150000018024 ms loss: -256.0
[compile] step: 31 step_time: 1.6974600000594364 ms loss: -256.0
[compile] step: 32 step_time: 1.6780400000016016 ms loss: -256.0
[compile] step: 33 step_time: 1.725410000062766 ms loss: -256.0
[compile] step: 34 step_time: 1.6667400000187627 ms loss: -256.0
[compile] step: 35 step_time: 1.725409999949079 ms loss: -256.0
[compile] step: 36 step_time: 1.7540300000291609 ms loss: -256.0
[compile] step: 37 step_time: 1.767489999906502 ms loss: -256.0
[compile] step: 38 step_time: 1.7174600000089413 ms loss: -256.0
[compile] step: 39 step_time: 1.7128199999660865 ms loss: -256.0
[compile] step: 40 step_time: 1.6362900000785885 ms loss: -256.0
[compile] step: 41 step_time: 1.7039800000020477 ms loss: -256.0
[compile] step: 42 step_time: 1.6502500000115106 ms loss: -256.0
[compile] step: 43 step_time: 1.6794800000070609 ms loss: -256.0
[compile] step: 44 step_time: 1.6123100000413615 ms loss: -256.0
[compile] step: 45 step_time: 1.5739300000632284 ms loss: -256.0
[compile] step: 46 step_time: 1.5075100000103703 ms loss: -256.0
[compile] step: 47 step_time: 1.5415200000461482 ms loss: -256.0
[compile] step: 48 step_time: 1.5012000000069747 ms loss: -256.0
[compile] step: 49 step_time: 1.5433800000437259 ms loss: -256.0
[compile] step: 50 step_time: 1.7087100000026112 ms
[compile] summary [16-40] total steps time: 41.28 ms, avg step time: 1.65 ms
pass_accuracy

欢迎加入社区，感谢您对社区的贡献 🎉!