Code

Benchmark#19

Code

Merged

jelmervdl merged 25 commits intomainjelmervdl/translatelocally-web-ext:mainfrom benchmarkjelmervdl/translatelocally-web-ext:benchmarkCopy head branch name to clipboardMay 6, 2022

+931-135

Commits

Commits on May 3, 2022

• WASM translator to accept options for native/embedded and number of workers

  jelmervdl

  committed
  1b22445

  Copy full SHA for 1b22445

• Add benchmark page

  jelmervdl

  committed
  95fbeea

  Copy full SHA for 95fbeea

• Less log spamming plz

  jelmervdl

  committed
  0b4a386

  Copy full SHA for 0b4a386

• Measure startup, quality of life improvements

  jelmervdl

  committed
  b494c9d

  Copy full SHA for b494c9d

Commits on May 4, 2022

• Add controls for lang/html config, add first translation timing for latency

  jelmervdl

  committed
  f13820a

  Copy full SHA for f13820a

• Add code to record page chunks during usage

  Show description for 82ccbc3
  jelmervdl

  committed
  82ccbc3

  Copy full SHA for 82ccbc3

• Allow for configuration of batch size in WASM translator

  jelmervdl

  committed
  e18790e

  Copy full SHA for e18790e

• More benchmarks, and make them toggleable

  jelmervdl

  committed
  83202fa

  Copy full SHA for 83202fa

• Make cache size configurable (and disable in benchmarks)

  jelmervdl

  committed
  f7b6d3b

  Copy full SHA for f7b6d3b

• Add Configure command for TranslateLocally

  jelmervdl

  committed
  50968c4

  Copy full SHA for 50968c4

• Merge branch 'main' into benchmark

  jelmervdl

  committed
  968542f

  Copy full SHA for 968542f

• Add test for cache size

  jelmervdl

  committed
  5085732

  Copy full SHA for 5085732

• Add little bar charts in the table background

  Show description for 821db60
  jelmervdl

  committed
  821db60

  Copy full SHA for 821db60

Commits on May 5, 2022

• Less "smart" implementation for lazy

  Show description for f2c94cc
  jelmervdl

  committed
  f2c94cc

  Copy full SHA for f2c94cc

• Fix untar issue in untar itself

  jelmervdl

  committed
  3263321

  Copy full SHA for 3263321

• Support multiple runs

  Show description for 229655a
  jelmervdl

  committed
  229655a

  Copy full SHA for 229655a

• Do the warm-up, but describe columns better

  jelmervdl

  committed
  cc03f8c

  Copy full SHA for cc03f8c

• batch-size 8 is default right now, makes more sense to test that one

  jelmervdl

  committed
  1d2eccd

  Copy full SHA for 1d2eccd

• Not using it but I like my Python asyncio.as_completed implementation

  jelmervdl

  committed
  cb1f7cf

  Copy full SHA for cb1f7cf

• Fix up popup

  jelmervdl

  committed
  fa0300d

  Copy full SHA for fa0300d

• Clean up record functionality

  jelmervdl

  committed
  0cc6897

  Copy full SHA for 0cc6897

• Hide all of this behind a "developer mode" toggle

  jelmervdl

  committed
  65a6d5d

  Copy full SHA for 65a6d5d

• Disable a bunch of benchmarks by default

  jelmervdl

  committed
  548621f

  Copy full SHA for 548621f

• Remove debug prints

  jelmervdl

  committed
  b6dbc9a

  Copy full SHA for b6dbc9a

• Remove the GEMM patching bit

  Show description for 88f3a98
  jelmervdl

  committed
  88f3a98

  Copy full SHA for 88f3a98