Skip to content

Latest commit

 

History

History
83 lines (51 loc) · 1.62 KB

CHANGELOG.md

File metadata and controls

83 lines (51 loc) · 1.62 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog.

This project adheres to Semantic Versioning, with the exception that v0.X updates include backwards-incompatible API changes. From v1.0.0 and on, the project will adherence strictly to Semantic Versioning.

[Unreleased]

[0.3.6] - 2023-08-08

Fixed

  • Updated on-gpu model benchmaking with best-practices on cuda.Event and cuda.synchronize.
  • FLOPs measurement error on CUDA.

[0.3.5] - 2023-06-09

Added

  • Repo DOI

[0.3.4] - 2022-02-22

Fixed

  • Add missing memory to results.

[0.3.3] - 2022-02-22

Fixed

  • Memory measurement for bs=1.

[0.3.2] - 2022-02-18

Fixed

  • Warm up batch size.

[0.3.1] - 2022-02-17

### Removed

  • try_custom_warmup.

Added

  • warm_up_fn overload option.
  • Support for FLOPs count in torch.nn.Module with input other than Tensor.

[0.3.0] - 2022-02-15

Added

  • Memory measurement for each batch size.
  • Repeated energy measurement.

Changed

  • Number formatting to use u instead of µ.

[0.2.2] - 2022-02-14

Added

  • Option to redirect info prints.

[0.2.1] - 2022-02-14

Fixed

  • Added missing with torch.no_grad

[0.2.0] - 2022-02-11

Added

  • Overloads for benchmark parameters and functions to allow benchmark of custom classes.

[0.1.2] - 2022-02-10

Fixed

  • GPU compatibility.

Removed

  • Carbon-tracker energy measurement. Library is still too immature at this point.

[0.1.1] - 2022-02-10

Added

  • Initial version.