v0.3.0

v0.3.0 - 2024-07-12

This release includes three major new features:

• Add support for Phi3 model.
• Add support for LoRA fine-tuning, which is specified via the new Tuning field in the CRD.
• Add support for using adapters for model inference, which is specified via the new "Adapters" field in the CRD.

Changelog

Features 🌈

• e5e61a7 feat: Add E2E Tests for Phi-3 and Tuning (#476)
• 42bbe92 feat: Bump Tags (#495)
• 766b944 feat: Add cloud provider specific nodeClaim requirement (#496)
• 2e06aff feat: Add Phi-3 Medium Plugin (#494)
• ada4b14 feat: Add Run Build Phi Flag (#493)
• 421bd5f feat: Add Phi-3 Manifests and Custom E2E Run Flag (#491)
• 001df06 feat: Add Target Modules Env Variable (#489)
• f534355 feat: Add Util funcs, updating func names, logs, configs and ensure service requirement (#485)
• f70dcc5 feat: Tuning Resource Validation Check (#484)
• 3ef8e66 feat: Add Phi3 Mini Requirements & Set Active Adapter (#469)
• d38d87b feat: Enhance CI/CD (Build, E2E, Composite Action) with 1ES Migration and Phi-3 Integration (#428)
• 4db58ce feat: Automate the adapters manifests (#463)
• 4cadcd2 feat: add phi-3 to configs.json
• 76a966b feat: Add Min GPU Memory Requirement (#443)
• e24e8b3 feat: Update Default Fine Tuning Params (with comments) (#442)
• 498f92b feat: Fine Tune (Part 10) - Updating fine tuning py (#371)
• 675a245 feat: Add karpenter labels to nodeClaim (#373)
• cb2d964 feat: Support Karpenter NodeClaim API in workspace (#366)
• 52f5d3b feat: Fine tune (Part 9) - Handling image data destination (#367)
• a4bcec4 feat: Add feature gates flag (#368)
• 536f259 feat: Add Karpenter Azure cloud provider NodePool (#369)
• a9af171 feat: Add URL as data source (#365)
• 962e3d6 feat: Add tuning job manifest, image source creation, parameter setup - Part 8 (#363)
• 2d5d1b4 feat: [SKU modularization] adding aws sku handler (#364)
• d9ba35a feat: Add v1beta1 API for Karpenter (#362)
• 5c3fb02 feat: [SKU modularization] adding azure sku handler (#360)
• aa89d20 feat: [SKU modularization] sku handler interface (#357)
• 0b63598 feat: Setup Preset Tuning Util Functions and miscellaneous validation/logging - Part 7 (#358)
• 6be8a0d feat: Simple Configmap Validation Checks - Part 6 (#355)
• bf4acba feat: Update CRD to add Volume and ConfigTemplate (#356)
• e799d62 feat: Add few util functions (#354)
• e2f5e25 feat: Add global client, accessible via webhooks (#353)
• f88b529 feat: Add default configmaps - Part 5 (#346)
• cf7cf94 feat: Add API Docs and Improve Inference Readability (#331)
• 05fb90c feat: Add sample front end helm chart (#320)
• 08dd1f4 feat: Initialize Fine-Tuning Interface and Core Methods - Part 3 (#308)
• 4ba337d feat: Part 2 - Add validation checks for TuningSpec, DataSource, DataDestination (#304)
• ec8a8e2 feat: Part 1 - Add FineTuning API (#201)
• 34f3266 feat: Add force update flag to e2e preset
• af64e89 feat: Add fine tuning spec in CRD (#292)
• 4fef28b feat: Add force run all flag for build images

Bug Fixes 🐞

• e4849a1 fix: Update Mistral Version and Rebuild (#501)
• d601ae8 fix: Bump plugins tags (#500)
• 9e12854 fix: Preset Build Image Naming (#497)
• 5abb030 fix: Remove Phi-3 Small and Update MCR Destination (#492)
• 21d7768 fix: Update e2e-preset-configs.json
• f2b3504 fix: Refactor Naming Conventions, Update Dependencies, Enhance Examples, and Add Volume Validation Check (#470)
• 8f068a3 fix: Adapter Base Image to Include Basic Utilities (#461)
• faa1fc7 fix: Standardize Dataset Input (#460)
• 261877e fix: Minor Tuning Results Filepath Bugfix (#453)
• 0dca449 fix: Update controller-gen version (#429)
• 4ae917b fix: Update dependabot.yml for security updates (#420)
• 64152d4 fix: Add preset nil checks (#347)
• ee9101a fix: Add registry as a pipeline job output (#329)
• d0498a0 fix: Upgrade FastAPI Version (#305)

Continuous Integration 💜

• cdcc7f5 ci: add setup-go GH action and codeQL analysis category (#401)
• 25a2b10 ci: bump codeQL to V3 (#400)
• 2aba9f0 ci: fix codeQL action version (#399)
• 799bda6 ci: fix codeql error due to golang version 1.22 (#397)
• 4bac047 ci: Add Env variable for supported models yaml (#376)
• c49b93a ci: Remove PR trigger from release workflow (#330)

Documentation 📘

• 174f6f5 docs: Add Tuning API Documentation (#480)
• e592eb3 docs: update instructions for llama2 (#475)
• 1812d4a docs: Update namespace patch and dependency for installation.md (#377)
• cc8e02d docs: Update gpu-provisioner installation step (#374)
• 2485c58 docs: Update README.md for announcing v0.2.1 (#307)

Maintenance 🔧

• 2cc97eb chore: remove deprecated goreleaser release actions (#499)
• 7b8416d chore: bump k8s.io/klog/v2 from 2.130.0 to 2.130.1 (#478)
• b4778e7 chore: bump github.com/samber/lo from 1.39.0 to 1.44.0 (#490)
• f613679 chore: e2e test for adapters and validation of adapters (#483)
• 5242169 chore: tuning e2e webhook test (#487)
• ce99b5f chore: bump codecov/codecov-action from 4.4.1 to 4.5.0 (#472)
• 45e75aa chore: bump actions/checkout from 4.1.6 to 4.1.7 (#473)
• 7b874eb chore: increase adapter test coverage (#477)
• d849274 chore: bump k8s.io/klog/v2 from 2.120.1 to 2.130.0 (#474)
• 5f03f40 chore: Update e2e-preset-configs.json
• 06fd57e chore: bump azure/login from 2.1.0 to 2.1.1 (#458)
• 42396db chore: bump goreleaser/goreleaser-action from 5.1.0 to 6.0.0 (#464)
• ffcafbb chore: bump step-security/harden-runner from 2.8.0 to 2.8.1 (#465)
• fb9b7da chore: bump actions/dependency-review-action from 4.3.2 to 4.3.3 (#466)
• 9968c29 chore: bump sigs.k8s.io/controller-runtime from 0.18.3 to 0.18.4 (#462)
• edbbf0e chore: bump docker/login-action from 3.1.0 to 3.2.0 (#457)
• 062ef5d chore: Copy latest CRD to chart (#444)
• 2888461 chore: update crd generated by 0.15 controller-gen (#441)
• 7cfb097 chore: bump sigs.k8s.io/controller-runtime from 0.18.2 to 0.18.3 (#433)
• 31d2f60 chore: bump github.com/onsi/ginkgo/v2 from 2.18.0 to 2.19.0 (#432)
• a2865ae chore: bump azure/login from 2.1.0 to 2.1.1 (#436)
• 6615a02 chore: bump step-security/harden-runner from 2.7.1 to 2.8.0 (#438)
• f3f4cbe chore: bump actions/dependency-review-action from 2.5.1 to 4.3.2 (#437)
• e5344e7 chore: bump github.com/go-logr/logr from 1.4.1 to 1.4.2 (#431)
• b9f2393 chore: bump github.com/onsi/ginkgo/v2 from 2.17.3 to 2.18.0 (#430)
• b65dd2e chore: bump sigs.k8s.io/karpenter from 0.36.1 to 0.36.2 (#426)
• 4a9b6c9 chore: bump step-security/harden-runner from 2.7.0 to 2.7.1 (#424)
• 92d8b71 chore: bump goreleaser/goreleaser-action from 5.0.0 to 5.1.0 (#422)
• a0eb42d chore: bump actions/checkout from 4.1.5 to 4.1.6 (#425)
• 060c91b chore: bump codecov/codecov-action from 4.3.1 to 4.4.1 (#427)
• 422fe20 chore: bump azure/login from 1.6.1 to 2.1.0 (#421)
• da3e63b chore: Update Kubernetes to 1.30.1 (#416)
• ff1d352 chore: bump datasets from 2.16.1 to 2.19.1 in /presets/tuning/tfs (#380)
• 8587d35 chore: Update Python Pip libraries (#412)
• 63fb951 chore: bump azure/login from 1.6.1 to 2.1.0 (#406)
• 19c93d6 chore: bump codecov/codecov-action from 4.1.0 to 4.3.1 (#405)
• 56e297b chore: bump actions/checkout from 3.6.0 to 4.1.5 (#404)
• f904cac chore: bump docker/login-action from 3.0.0 to 3.1.0 (#407)
• 9fe0eef chore: bump azure/CLI from 1.0.9 to 2.0.0 (#408)
• 2c05921 chore: update ginkgo version for e2e test (#403)
• 2f6e1d9 chore: Remove Karpenter noodepool (#402)
• 63f9f72 chore: bump github.com/stretchr/testify from 1.8.4 to 1.9.0 (#393)
• 8fabf20 chore: bump distroless/static from 55c6361 to e9ac71e in /docker/kaito (#388)
• 37acb67 chore: bump github.com/onsi/ginkgo/v2 from 2.17.1 to 2.17.3 (#394)
• 2c42bdd chore: remove gpu-provisioner chart from kaito repo (#372)
• 8b7941a chore: Update Golong 1.20 -> 1.22 (#361)
• 65a99da chore: naming nit preset-inferences.go
• 9e2ac24 chore: Add utils functions and move tests into seperate package (#351)
• 77aa95b chore: Factoring out reusable presets logic - Part 4 (#332)
• 2f0323a chore: Add GitHub issue/PR templates (#316)

Security Fix 🛡️

• 667d426 security: Apply security best practices (#318)

Testing 💚

• 9342647 test: Add UTs for invalid image source or destination (#359)