Navigation

    全志在线开发者论坛

    • Register
    • Login
    • Search
    • Categories
    • Tags
    • 在线文档
    • 社区主页

    关于量化时添加 --MLE 参数

    V Series
    2
    3
    2354
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • C
      cruise33 LV 6 last edited by

      pegasus quantize --model yolo.json --model-data yuntai.data --MLE --algorithm kl_divergence --batch-size 1 --iterations 1 --divergence-nbins 4096 --device CPU --with-input-meta ./yolo_inputmeta.yml --rebuild --model-quantize yolo.quantize --quantizer asymmetric_affine --qtype uint8 
      

      怎么加了--MLE(最小化每层误差)参数后,量化的时候超级久,batch-size为1时差不多要2个小时, 再设大点好几天都没量化完,哪位大神了解嘛?

      WhycanService 1 Reply Last reply Reply Quote Share 0
      • WhycanService
        WhycanService LV 8 @cruise33 last edited by

        @cruise33 MLE 最好用带GPU的量化,这东西算力需求有点大

        C 1 Reply Last reply Reply Quote Share 0
        • C
          cruise33 LV 6 @WhycanService last edited by

          @whycanservice 量化工具好像没有适配cuda,用不了GPU

          1 Reply Last reply Reply Quote Share 0
          • 1 / 1
          • First post
            Last post

          Copyright © 2024 深圳全志在线有限公司 粤ICP备2021084185号 粤公网安备44030502007680号

          行为准则 | 用户协议 | 隐私权政策