Skip to content

Latest commit

 

History

History
151 lines (115 loc) · 4.26 KB

Releasenotes.md

File metadata and controls

151 lines (115 loc) · 4.26 KB
Raw

Version 1.9.6

  • Fixed a bug where trainer couldn't get a task (object reference not set...)
  • Re-Added CPU autodetection for new trainer config scheme (Auto AVX2/AVX512)

Version 1.9.5

  • Added GPU Cards Feature (select which GPU Cards to use and how many threads or let the client auto tune them)

Version 1.9.4

  • Minor Bugfixes

Version 1.9.3

  • Native AMD GPU support
  • Added new options to select your trainer
  • Prepared for combined CPU+GPU training

Version 1.9.2

skipped version

Version 1.9.1

  • fixed windows trainer change problem

Version 1.9.0

  • optimized trainer handling
  • added possibility to configure cpu afinity
  • internal optimizations

Version 1.8.10

  • fixed cases when the client crashed unexpected

Version 1.8.9

  • changed internal solution processing
  • added auto re-register of client if registration the first time fails (lead to 'waiting for task' loop)
  • fixed some minor bugs

Version 1.8.8

  • added ability to resend unsubmitted solution after a client restart

Version 1.8.7

  • Log time set to UTC
  • added graceful handling of runner
  • update to .net 8
  • optimized logging of found solutions during epoch

Version 1.8.6

  • Fixed wrong "skylake" overwrite for linux client

Version 1.8.5

  • Optimized Epoch Change

Version 1.8.4

  • Added new Output line for GPU Trainers GPU #0: 870 it/s | GPU #1: 884 it/s | GPU #2: 870 it/s | GPU #3: 872 it/s to have this information available in hiveos

Version 1.8.3

  • Fixed a bug in calculating local found solutions. (e.g. wrong "SOL: 12/300" display)
  • Fiexd a reporting bug
  • Added experimental display of epoch to console output

Version 1.8.2

  • Fixed a bug in writing stats.EPOCH.lock file

Version 1.8.1

  • Added optimized console logging. (new format: "SOL: X/Y" where Y is found solutions and X is submitted solutions)

Version 1.8.0

  • GPU Release

Version 1.7.6

  • Fixed Bug where solution finding was delayed
  • Changed main communication Scheme from qiner to qli internal

Version 1.7.5

  • Windows Release
  • Bugfixes

Version 1.7.3

  • UPgrade from .Net 6 to 7
  • Added Intel Skylake detection for Linux

Version 1.7.2

  • Downgrade from .Net 7 to 6
  • Removed AVX2 Fallback for Linux

Version 1.7.1

  • Optimized Network Traffic Handling
  • Significant performance boost

Version 1.7.0

  • Fixed Bug in Solution Delivery in certain environments
  • Added compatibility for new Qubic protocol

Version 1.6.1

  • Fixed potential Bug in Solution Delivery
  • Added http/2 support

Version 1.6.0

  • Improved API Handling
  • General Bugfixing
  • Added/Updated Qiner Protocol

Version 1.5.5

  • fixed bug on Windows with stopping to report status
  • automatic fallback to AVX2 on linux
  • DEPRECATED: custom QLI Protocol

Version 1.5.4

  • fixed bug with Qiner communication protocol

Version 1.5.2

  • fixed bug which decreases training performance over time

Version 1.5.2

  • Optimized Communication between Client and API
  • Fixed blocking Bug
  • Deprecated: "AVX2":true setting
  • Optimized Detection of AVX512 and added Fallback to AVX2
  • Removed Deprecated Settings
  • Added Setting AllowOptimization to allow the trainer to try to optimize environment. (Default: false)
  • Reduced Runner Footprint

Version 1.4.2

  • Bugfixing
  • Automatic optimization of the environment

Version 1.4.0

  • Bugfixing
  • Possibility to Overwrite Params like AVX512
  • new argument --version to get current version of Client

Version 1.3.9

  • Bugfixing
  • New Setting: "serviceLock": true to use qiner protocol for custom runner
  • New Log Format with display of average it/s
  • Signaling if the trainer isn't starting properly

Version 1.3.5

  • Fixed AVX512 detection on Windows
  • Fixed Thread Settings from Portal
  • Automatic runner selection from Portal Settings
  • Automatic thread selection from Portal Settings

Version 1.3.4

  • New Setting: "useAvx": true which forces the client to use AVX2 runner
  • New Setting: "allowHwInfoCollect": true which will allow the client to collect CPU Model, RAM Size and CPU Cache Size; with this information the client can better decide which runner should be used
  • Optimized Detection of AVX2/512
  • Optimized Error Handling

Version 1.3.0

  • Added automated selection of AVX2/512 runner

Version 1.2.3