YingkunZhou comments

Results 13 comments of


                                            YingkunZhou

question about Firrtl(IR) optimization

But you cannot expect that synthesis tools will do these things.

question about Firrtl(IR) optimization

;buildInfoPackage: chisel3, version: 3.1.5, scalaVersion: 2.11.12, sbtVersion: 1.1.1, builtAtString: 2018-12-14 23:45:23.572, builtAtMillis: 1544831123572 circuit Adder : module Adder : input clock : Clock input reset : UInt output io :...

question about Firrtl(IR) optimization

not familiar with firrtl, though

question about Firrtl(IR) optimization

so you use master branch? I forget when and how I installed firrtl

INT8 量化模型在 ARM Android上的性能严重缩水，无法复现在benchmark.md中的加速比

您好，就是默认分支的最新commit: 3c61295edf66eb5c24e7c98b604c14a1bd4457d7

quantized.out int8量化后的模型在CPU后端运行resnet50出错（已更新到最新commit版本）

继续用经典的模型efficientnetv2_b0试了试模型的下载地址为：https://github.com/onnx/models/raw/main/Computer_Vision/tf_efficientnetv2_b0_Opset16_timm/tf_efficientnetv2_b0_Opset16.onnx 得到的量化完的运行结果： ```bash ./pictureRecognition.out tf_efficientnetv2_b0_Opset16.mnn daisy.jpg Load Cache file error. The device support i8sdot:1, support fp16:1, support i8mm: 0 Create execution error : 101 Create execution error : 101...