Releases: nutanix/nai-llm-k8s
Releases · nutanix/nai-llm-k8s
v0.2.2
What's Changed
- pin setuptools to v69.5.1 in model_requirements.txt by @lauranutanix in #54
- Update download logic by @saileshd1402 in #55
New Contributors
- @lauranutanix made their first contribution in #54
Full Changelog: v0.2.1...v0.2.2
v0.2.2-rc1
What's Changed
- pin setuptools to v69.5.1 in model_requirements.txt by @lauranutanix in #54
- Update download logic by @saileshd1402 in #55
New Contributors
- @lauranutanix made their first contribution in #54
Full Changelog: v0.2.1...v0.2.2-rc2
v0.2.1
This is the stable release of NAI-LLM-K8s v0.2.1
What's Changed
- Chatbot demo for K8s version by @AyushSawant18588 in #48
- Quantization Support by @saileshd1402 in #46
- HF Token Fixes by @saileshd1402 in #47
- Bug fix and Transformer version upgrade by @gavrishp in #51
Full Changelog: v0.2...v0.2.1
v0.2.1-rc2
What's Changed
- Refresh request map for every process by @gavrishp in #51
- Chatbot demo for K8s version by @AyushSawant18588 in #48
- Quantization Support by @saileshd1402 in #46
- HF Token Fixes by @saileshd1402 in #47
Full Changelog: v0.2...v0.2.1-rc2
v0.2.1-rc1
What's Changed
- Chatbot demo for K8s version by @AyushSawant18588 in #48
- Quantization Support by @saileshd1402 in #46
- HF Token Fixes by @saileshd1402 in #47
Full Changelog: v0.2...v0.2.1-rc-1
v0.2
This is the stable release of NAI-LLM-K8s v0.2
Features
- Custom HuggingFace Models support
Additional models supported
- codellama_7b_python
- llama2_7b_chat
What's Changed
- Download support for custom Huggingface models by @AyushSawant18588 in #40
- Name change of download file and no download flag by @AyushSawant18588 in #41
- Add license Info by @johnugeorge in #42
- Added support for llama2-chat and codellama models by @AyushSawant18588 in #43
- Updated transformers (4.36.0) and tokenizers (0.15.0) version by @AyushSawant18588 in #44
Important notes
- V2 does not have backward compatibility with V1. Follow new documentation in opendocs
New Contributors
- @johnugeorge made their first contribution in #42
Full Changelog: v0.1...v0.2
v0.2-rc2
What's Changed
- Download support for custom Huggingface models by @AyushSawant18588 in #40
- Name change of download file and no download flag by @AyushSawant18588 in #41
- Add license Info by @johnugeorge in #42
- Added support for llama2-chat and codellama models by @AyushSawant18588 in #43
- Updated transformers (4.36.0) and tokenizers (0.15.0) version by @AyushSawant18588 in #44
Important notes
- V2 does not have backward compatibility with V1. Follow new documentation in opendocs
New Contributors
- @johnugeorge made their first contribution in #42
Full Changelog: v0.1...v0.2-rc2
v0.2-rc1
What's Changed
- Download support for custom Huggingface models by @AyushSawant18588 in #40
- Name change of download file and no download flag by @AyushSawant18588 in #41
- Add license Info by @johnugeorge in #42
Important notes
- V2 does not have backward compatibility with V1. Follow new documentation in opendocs
New Contributors
- @johnugeorge made their first contribution in #42
Full Changelog: v0.1...v0.2-rc1
v0.1
Introduce NAI-LLM run support on k8s
Models Support
- MPT_7b
- Falcon_7b
- Llama2_7b
- Phi-1_5
- GPT2
Additional Features
- Custom trained model support
- Model version based support
Repo Stability
- Linter test included in Actions with Black and PyLint tests
- Product stability Test Sets created with PyTest
Changes
- Kubeflow inference run support for LLMs by @gavrishp in #2
- Introduce support for deployment cleanup by @AyushSawant18588 in #8
- Model repository version support added by @AyushSawant18588 in #18
- Health check support and make data param optional by @gavrishp in #19
- Introduce support to gpt2 model by @gavrishp in #21
- Add Python Linters by @AyushSawant18588 in #23
- Bump kserve version to 0.11.1 by @gavrishp in #20
- Custom mar support by @AyushSawant18588 in #30
- Pytest Github Action Integration by @AyushSawant18588 in #32
New Contributors
- @nagar-ajay made their first contribution in #1
- @gavrishp made their first contribution in #2
- @AyushSawant18588 made their first contribution in #8
- @saileshd1402 made their first contribution in #29
Full Changelog: https://github.com/nutanix/nai-llm-k8s/commits/v0.1
v0.1-rc5
This is the rc5 of NAI-LLM-k8s v0.1
Include support for MPT_7b, Falcon_7b, Llama2_7b, Phi-1_5 and GPT2
What's Changed
- Update readme and kfp_version variable name by @nagar-ajay in #1
- Kubeflow inference run support for LLMs by @gavrishp in #2
- fix path in install script by @nagar-ajay in #3
- include support for Model params by @gavrishp in #4
- Update inference port in config properties by @gavrishp in #9
- Introduce support for deployment cleanup by @AyushSawant18588 in #8
- Model repository version support added by @AyushSawant18588 in #18
- Health check support and make data param optional by @gavrishp in #19
- Introduce support to gpt2 model by @gavrishp in #21
- Add Python Linters by @AyushSawant18588 in #23
- Bump kserve version to 0.11.1 by @gavrishp in #20
Changes from v0.1-rc4
- Custom mar support by @AyushSawant18588 in #30
- Pytest Github Action Integration by @AyushSawant18588 in #32
Changelog from rc4: v0.1-rc4...v0.1-rc5
New Contributors
- @nagar-ajay made their first contribution in #1
- @gavrishp made their first contribution in #2
- @AyushSawant18588 made their first contribution in #8
- @saileshd1402 made their first contribution in #29
Full Changelog: https://github.com/nutanix/nai-llm-k8s/commits/v0.1-rc5