v0.1.2.3

oumi-ai/oumiv0.1.2.3Jan 27, 2025by oelachqar

AI Summary

This release focuses on bug fixes, including improvements to GGUF support and pyright compatibility, along with various documentation updates.

Key Highlights

  • Improved support for GGUF models (incl. auto-download)
  • Fixed pyright breakage when vllm and llama_cpp are not installed
  • Added e2e evaluation tests
  • Updated documentation and quickstart guides

New Features

  • VLLM Engine improvements for GGUF
  • Documentation updates for evaluation and judge
  • Updated workflow names for PyPI release

Full Release Notes

## What's Changed
* Re-enable parallel evaluation for VLM-s by @xrdaukar in https://github.com/oumi-ai/oumi/pull/1235
* Add multimodal exemplar dataset in our provided mini-datasets by @optas in https://github.com/oumi-ai/oumi/pull/1234
* [Tiny] renaming a field name (`init_lora_weights`) by @optas in https://github.com/oumi-ai/oumi/pull/1236
* Add more e2e evaluation tests by @xrdaukar in https://github.com/oumi-ai/oumi/pull/1237
* Fix pyright breakage when vllm and llama_cpp are not installed by @taenin in https://github.com/oumi-ai/oumi/pull/1240
* Update our oumi launch documentation. by @taenin in https://github.com/oumi-ai/oumi/pull/1239
* Update index.md title for "Join the Community!" by @mkoukoumidis in https://github.com/oumi-ai/oumi/pull/1242
* Update quickstart.md - nit for Oumi support request by @mkoukoumidis in https://github.com/oumi-ai/oumi/pull/1241
* [VLLM Engine] Improve support for GGUF models (incl. auto-download) by @kaisopos in https://github.com/oumi-ai/oumi/pull/1238
* Update README.md title to "Join the Community!" by @mkoukoumidis in https://github.com/oumi-ai/oumi/pull/1243
* Update quickstart.md by @brragorn in https://github.com/oumi-ai/oumi/pull/1251
* Update quickstart.md by @brragorn in https://github.com/oumi-ai/oumi/pull/1253
* Update quickstart.md by @brragorn in https://github.com/oumi-ai/oumi/pull/1252
* Update quickstart.md by @brragorn in https://github.com/oumi-ai/oumi/pull/1250
* [Minor refactor] Moving model caching to `oumi.utils` by @kaisopos in https://github.com/oumi-ai/oumi/pull/1246
* Add more details to troubleshooting FAQ by @wizeng23 in https://github.com/oumi-ai/oumi/pull/1249
* Update training_methods.md - Change compute requirement suggestions by @mkoukoumidis in https://github.com/oumi-ai/oumi/pull/1245
* Update train.md - nit description change by @mkoukoumidis in https://github.com/oumi-ai/oumi/pull/1244
* [docs] misc docs feedback by @oelachqar in https://github.com/oumi-ai/oumi/pull/1248
* [tiny] Qwen2-VL activate experimental datapipes by @optas in https://github.com/oumi-ai/oumi/pull/1247
* Update Oumi - A Tour.ipynb by @brragorn in https://github.com/oumi-ai/oumi/pull/1254
* [docs] more docs feedback by @oelachqar in https://github.com/oumi-ai/oumi/pull/1255
* Update supported_models.md by @penfever in https://github.com/oumi-ai/oumi/pull/1256
* Rename `experimental_use_torch_datapipes` data param by @xrdaukar in https://github.com/oumi-ai/oumi/pull/1257
* Add pypi release workflow using testpypi by @oelachqar in https://github.com/oumi-ai/oumi/pull/1259
* Update workflow names by @oelachqar in https://github.com/oumi-ai/oumi/pull/1262
* Update default idle_minutes_to_autostop to 1 hour. by @taenin in https://github.com/oumi-ai/oumi/pull/1264
* update pypi release workflow to use trusted env by @oelachqar in https://github.com/oumi-ai/oumi/pull/1265


**Full Changelog**: https://github.com/oumi-ai/oumi/compare/v0.1.2.2...v0.1.2.3