Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit history

Commits on Jun 16, 2026

• finalize logging

  Show description for 533db92
  mrelmida

  committed
  533db92

  Copy full SHA for 533db92

Commits on Jun 14, 2026

• Merge branch 'SearchSavior:main' into logging-update

  mrelmida

  authored
  5fd8646

  Copy full SHA for 5fd8646

Commits on May 25, 2026

• Make cache path configurable per-model, and support all models

  Show description for ed235d2
  solidDoWant

  committed
  ed235d2

  Copy full SHA for ed235d2

• Merge pull request #123 from SearchSavior/dont-require-vlm-type

  Show description for e3618c0
  SearchSavior

  authored
  e3618c0

  Copy full SHA for e3618c0

Commits on May 23, 2026

• Fix bugs with relative model path resolution

  Show description for 528ae1f
  solidDoWant

  committed
  528ae1f

  Copy full SHA for 528ae1f

• implement and document transformers config.json "architectures" key parsing to replace vlm_type

  SearchSavior

  committed
  b7f0940

  Copy full SHA for b7f0940

Commits on May 9, 2026

• support saving device and runtime_config to db during benchmark

  Show description for 9cdffbf
  SearchSavior

  committed
  9cdffbf

  Copy full SHA for 9cdffbf

Commits on May 7, 2026

• Add config option for Qwen3.5 VLM type

  droans

  committed
  66299c1

  Copy full SHA for 66299c1

• introduce openarc list prunem which will remove stale config entries

  

  SearchSavior

  authored and

  droans

  committed
  8acc9dc

  Copy full SHA for 8acc9dc

• - added tool to inspect openivno ir files for the ops used in their graph

  

  SearchSavior

  authored and

  droans

  committed
  82d15bd

  Copy full SHA for 82d15bd

• - Support Gemma4 ! Wooohoooo

  

  SearchSavior

  authored and

  droans

  committed
  3720ef8

  Copy full SHA for 3720ef8

• - remove reporting p_actual; accept that adding chat_template tokens is fine for now

  

  SearchSavior

  authored and

  droans

  committed
  28ac552

  Copy full SHA for 28ac552

• - initial commit of vlm tokenizer consistency workaround to enable benchmarking VLMPipeline

  

  SearchSavior

  authored and

  droans

  committed
  ac99623

  Copy full SHA for ac99623

• - update API key requirements.

  

  SearchSavior

  authored and

  droans

  committed
  28dbe7c

  Copy full SHA for 28dbe7c

• - support qwen3-vl

  

  SearchSavior

  authored and

  droans

  committed
  3db98d9

  Copy full SHA for 3db98d9

Commits on May 4, 2026

• added verbosity setting and fixed a index naming in gpu_metrics

  mrelmida

  committed
  a14e86e

  Copy full SHA for a14e86e

Commits on May 1, 2026

• - Support Gemma4 ! Wooohoooo

  SearchSavior

  committed
  023e79c

  Copy full SHA for 023e79c

Commits on Apr 17, 2026

• - remove reporting p_actual; accept that adding chat_template tokens is fine for now

  SearchSavior

  committed
  16beac1

  Copy full SHA for 16beac1

Commits on Apr 16, 2026

• - initial commit of vlm tokenizer consistency workaround to enable benchmarking VLMPipeline

  SearchSavior

  committed
  107b0a8

  Copy full SHA for 107b0a8

Commits on Apr 15, 2026

• - update API key requirements.

  SearchSavior

  committed
  77ff5dc

  Copy full SHA for 77ff5dc

Commits on Apr 11, 2026

• - support qwen3-vl

  SearchSavior

  committed
  ae5c7f7

  Copy full SHA for ae5c7f7

Commits on Apr 1, 2026

• - add -depth, to simulate context in openarc bench

  SearchSavior

  committed
  60b283e

  Copy full SHA for 60b283e

Commits on Mar 14, 2026

• - qwen3 tts model load, unload, queues, pydantic models

  SearchSavior

  committed
  cdaa300

  Copy full SHA for cdaa300

Commits on Feb 18, 2026

• - initial commit of qwen3_asr

  SearchSavior

  committed
  d0e278d

  Copy full SHA for d0e278d

Commits on Feb 11, 2026

• - full refactor of command line tool to make adding more advanced features possible.

  Show description for cfd1a9c
  SearchSavior

  committed
  cfd1a9c

  Copy full SHA for cfd1a9c

Loading