Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Jan 12, 2023

• Update benchmark.py

  justinpinkney

  authored
  df83334

  Copy full SHA for df83334

Commits on Dec 20, 2022

• Update benchmark script, add Dockerfile

  justinpinkney

  authored
  cdddab8

  Copy full SHA for cdddab8

Commits on Oct 11, 2022

• remove erroneous import

  eolecvk

  committed
  5cf05f4

  Copy full SHA for 5cf05f4

Commits on Oct 7, 2022

• cuda autocast y/n argument

  eolecvk

  committed
  2c1e19e

  Copy full SHA for 2c1e19e

Commits on Oct 5, 2022

• missing csv header 'runtime'

  eolecvk

  committed
  2b5da79

  Copy full SHA for 2b5da79

• saving logs to benchmark_tmp.csv instead of benchmark.csv and gitignore _tmp.csv; better try/catch if missing ACCESS TOKEN

  eolecvk

  committed
  645fc33

  Copy full SHA for 645fc33

Commits on Oct 3, 2022

• benchmark: Add readme to benchmark

  chuanli11

  committed
  cc81843

  Copy full SHA for cc81843

• benchmark: add revsiion when loading pretrianed models for half precision benchmark

  chuanli11

  committed
  2469d61

  Copy full SHA for 2469d61

Commits on Oct 2, 2022

• benchmark: use autocast in a more elegant way. remove onnx from gpu benchmark

  chuanli11

  committed
  37b5afd

  Copy full SHA for 37b5afd

• benchmark: handles onnx runtime error

  chuanli11

  committed
  254536f

  Copy full SHA for 254536f

Commits on Oct 1, 2022

• benchmark: use float instead of str for logging

  chuanli11

  committed
  e873162

  Copy full SHA for e873162

• benchmark: 1) Catch CUDA OOM error more gracefully 2) Add samples (batch size) as a cmd line argument.

  chuanli11

  committed
  91542b2

  Copy full SHA for 91542b2

• benchmark: 1) Add onnx 2) Add arguments to benchmark script

  chuanli11

  committed
  e86d69e

  Copy full SHA for e86d69e

• benchmark: 1) Get precisons working 2) Add cpu support

  chuanli11

  committed
  656c4d1

  Copy full SHA for 656c4d1

Commits on Sep 29, 2022

• csvpath relative to benchmark script

  eolecvk

  committed
  f54cfbb

  Copy full SHA for f54cfbb

• append to benchmark, add device col

  eolecvk

  committed
  8cf9a0b

  Copy full SHA for 8cf9a0b

• load model from huggingface

  eolecvk

  committed
  90bf85f

  Copy full SHA for 90bf85f

Commits on Sep 28, 2022

• load weights from repo, add n_samples option, run benchmark grid, save to tidy csv

  eolecvk

  committed
  3eee230

  Copy full SHA for 3eee230

• rename benchmark script, add memory log

  eolecvk

  committed
  136903a

  Copy full SHA for 136903a

Commits on Sep 27, 2022

• switch from timeit to torch.utils.benchmark.Timer.timeit

  eolecvk

  committed
  0673aff

  Copy full SHA for 0673aff

Commits on Sep 26, 2022

• log inference time

  eolecvk

  committed
  b590b3f

  Copy full SHA for b590b3f

Commits on Sep 12, 2022

• turn into package

  justinpinkney

  committed
  5d3d440

  Copy full SHA for 5d3d440

Loading