v0.6.5 🐋

🧭 What's Changed

• bugfix(storage): add ksyun configuration (#2987) @xuchuan
• chore(ci): reduce ut log output of controller (#2963) @jialeicui
• chore(controller): add dataset info for step spec (#2972) @goldenxinxing
• chore(controller): add draft field to model version vo (#2982) @anda-ren
• chore(controller): add more default values in system setting (#2985) @anda-ren
• chore(controller): enable scheduler event log by default (#2937) @jialeicui
• chore(controller): keep release api align with UI (#2981) @anda-ren
• chore(controller): prevent collecting duplicate scheduler events after controller restart (#2954) @jialeicui
• chore(controller): support disabling the fine tuning feature (#2975) @jialeicui
• chore(controller): validate the resource version owner (#2934) @jialeicui
• chore(datastore): shorten first item response time (#2940) @jialeicui
• chore(example): add finetune example for stable diffusion (#2641) @goldenxinxing
• chore(sdk): revert "enhance(SDK): enhance hf datasets to handle abnormal dataset without config class" (#2947) @tianweidut
• chore: image sync loss star-whale (#2977) @goldenxinxing
• enhance(SDK): enhance hf datasets to handle abnormal dataset without config class (#2945) @tianweidut
• enhance(console): show running serving pages of current model version (#2942) @jialeicui
• example(runtime): update pytorch example runtime (#2984) @tianweidut
• example: add evaluate phase for segment anything model (#2938) @tianweidut
• example: add image classification example (#2941) @tianweidut
• example: add vit example for image classification (#2944) @tianweidut
• example: support segment anything model (#2929) @tianweidut
• feat(client): support finetune model source field for model copy (#2967) @tianweidut
• feat(console): support job event with filter/download/fullscreen etc (#2935) @waynelwz
• feat(console): support project readme (#2936) @waynelwz
• feat(controller): Sft/model copy with ft id (#2966) @anda-ren
• feat(controller): add CRUD for finetune and space (#2950) @anda-ren
• feat(controller): add SFT release feature (#2976) @anda-ren
• feat(controller): add ft eval create job api (#2971) @goldenxinxing
• feat(controller): add model source when copy (#2953) (#2964) @anda-ren
• feat(controller): add sft info api (#2983) @anda-ren
• feat(controller): fill SFT list fields (#2980) @anda-ren
• feat(controller): support more file type for dataset build (#2957) @goldenxinxing
• feat(sdk): finetune decorator supports auto build and validation datasets (#2962) @tianweidut
• feat(storage): support ksyun (#2986) @xuchuan
• fix(client): fix broken resource ls cmd (#2969) @jialeicui
• fix(console): action bar not rerender (#2988) @waynelwz
• fix(console): eval action position when scrolling (#2961) @waynelwz
• fix(console): overview link/project desc overflow/tree default selector (#2952) @waynelwz
• fix(controlle): remove unique constrain for job on finetune (#2970) @anda-ren
• fix(controller): fix compile error (#2965) @anda-ren
• fix(controller): fix desc not null (#2968) @anda-ren
• fix(controller): fix sql error in sft release function (#2990) @anda-ren
• fix(controller): npe in ds build job (#2956) @goldenxinxing
• fix(controller): project update to private when pass null by api (#2943) @goldenxinxing
• fix(controller): train and eval dataset wrong (#2989) @goldenxinxing
• fix(datastore): fix NPE when list diff without sparse element type inited (#2974) @jialeicui
• optimie(console): new action bar for first class list pages (#2949) @waynelwz
• refactor(console): user auto generated models (#2960) @jialeicui
• refactor(controller): deprecate resource version url in job request (#2939) @jialeicui
• refactor(controller): remove unreachable code (#2951) @jialeicui
• update(console): use swagger template api and schema as default api (#2955) @waynelwz

⚙️ Who Contributes

@anda-ren, @goldenxinxing, @jialeicui, @tianweidut, @waynelwz and @xuchuan

v0.6.4 🐋

🧭 What's Changed

• chore(client): list tables without switch instance in datastore debug cli (#2931) @jialeicui
• chore(client): remove redunancy atexit for data_store (#2932) @tianweidut
• enhance(client): prefer using the built-in map when building dataset (#2927) @jialeicui
• enhance(console): support embedded online eval page (#2928) @jialeicui
• example: datasets for image segmentation (#2925) @tianweidut
• feat(controller): add overview for project (#2923) @goldenxinxing
• fix(datastore): prevent npe while deal with map diff (#2930) @jialeicui
• refactor(controller): make read log generate to async and add scanRange impl (#2907) @goldenxinxing

⚙️ Who Contributes

@goldenxinxing, @jialeicui and @tianweidut

v0.6.3 🐋

🧭 What's Changed

• bug(runtime): fix runtime.yaml arch field empty issue (#2917) @tianweidut
• chore(controller): add missing related resource field in job event (#2922) @jialeicui
• chore(controller): list event support task-scoped aggregation (#2915) @jialeicui
• chore(db): change desc column's character set to utf8mb4 (#2920) @goldenxinxing
• enhance(client): online serving handler support the latest gradio (#2924) @jialeicui
• enhance(client): support login with browser (#2916) @jialeicui
• enhance(datastore): map support multiple key value schemas (#2873) @jialeicui
• example: add code-llama example (#2911) @tianweidut
• fix(client): model serving handler can not run (#2919) @jialeicui
• fix(controller): add missing owner for dataset and model (#2926) @jialeicui
• fix(controller): make proxy works after serving handler run (#2921) @jialeicui

⚙️ Who Contributes

@goldenxinxing, @jialeicui and @tianweidut

v0.6.2 🐋

🧭 What's Changed

• bug(client): enqueue None to the abs queue after artifact close operation (#2894) @tianweidut
• chore(ci): fix evaluation test typo (#2879) @tianweidut
• chore(ci): improve tracker run test stability in the slow machine (#2876) @tianweidut
• chore(client): auto generated model uses custom base model (#2910) @jialeicui
• chore(client): enable flake8 bugbear for python code (#2897) @tianweidut
• chore(client): make unit test works after fastapi upgrade (#2881) @jialeicui
• chore(client): prefer using builtin list when building dataset (#2913) @jialeicui
• chore(client): prevent showing the 🆔 emoji in console (#2871) @jialeicui
• chore(client): replace flake8 with ruff, an extremely fast Python linter (#2896) @tianweidut
• chore(client): support project opt when listing table in debug cli (#2867) @jialeicui
• chore(client): workaround email-validator lib version in python3.7 (#2892) @tianweidut
• chore(controller): datastore use read-write lock (#2895) @goldenxinxing
• chore(controller): ignore k8s job fail reason and refine task finish time (#2912) @jialeicui
• chore(controller): use try-resource to recode pagehelper (#2893) @goldenxinxing
• chore(e2e): add jenkins e2e test case example (#2898) @anda-ren
• chore(e2e): compare datastore table instead of dataset (#2902) @jialeicui
• chore: speed up image sync (#2903) @anda-ren
• chore: update docs for sdk and example (#2874) @tianweidut
• ci(e2e): add comparision between local and remote dataset (#2861) @jialeicui
• enhance(SDK): avoid obtaining all artifact content in advance (#2882) @tianweidut
• enhance(SDK): retrieve the content of Text using the fetch_data function (#2891) @tianweidut
• enhance(ci): auto cancel the previous run (#2906) @jialeicui
• enhance(client): clear cache for link _signed_uri (#2880) @tianweidut
• enhance(client): remove clear_cache operation for data types (#2888) @tianweidut
• enhance(client): remove artifacts owner setter (#2886) @tianweidut
• enhance(controller): add job event in server side (#2900) @jialeicui
• enhance(controller): event api support run id (#2877) @jialeicui
• enhance(controller): modify datasource param and add index for slowsql (#2872) @goldenxinxing
• enhance(controller): support logging scheduler event to run of job (#2904) @jialeicui
• enhance(datastore): support multiple type of list (#2838) @jialeicui
• example: add musicgen example (#2899) @tianweidut
• feat(console): grid table support pin right and add action column to eval list (#2901) @waynelwz
• feat(console): support model/project quick notice with locale (#2884) @waynelwz
• fix(client): make sure evaluation log flushed to storage (#2875) @jialeicui
• fix(controller): add start and stop time for run when canceled (#2868) @anda-ren
• fix(controller): failed task that is resumed have two runs (#2869) @anda-ren
• fix(controller): fix jobs that run on k8s are always cancelling when cancel operation is commited (#2864) @anda-ren
• fix(controller): make online eval runable (#2914) @jialeicui
• refactor(controller): unify sign link api for dataset and evaluation (#2887) @anda-ren
• refine(client): dataset head always fetch raw (#2870) @jialeicui
• update(console): action configable, report autofocus, job result overflow scroll and optimize ui (#2909) @waynelwz
• update(console): add handler to model overview (#2889) @waynelwz
• update(console): eval bar support datepicker with range/time/calender selector (#2885) @waynelwz
• update(console): use current project as default tree data list (#2905) @waynelwz

⚙️ Who Contributes

@anda-ren, @goldenxinxing, @jialeicui, @tianweidut and @waynelwz

v0.6.1 🐋

🧭 What's Changed

• feat(controller): add run info into taskVo (#2853) @anda-ren
• fix(controller): fix when explicit_defaults_for_timestamp is off, run table could not be created (#2865) @anda-ren
• fix(datastore): normalize the revision in snapshot for backward compatibility (#2866) @jialeicui

⚙️ Who Contributes

@anda-ren and @jialeicui

v0.6.0 🐋

🍭 Release Highlights

Starwhale 0.6.0 release brings essential performance and stability improvements for the datastore component. At the same time, Python SDK, dataset, and UI have also received some critical updates.

• Datastore: In this release, we have enhanced the performance of the WAL log dumping process (#2782, #2800, #2805) and addressed the issue of unexpected empty data being automatically added in specific scenarios (#2825) on the Server side. On the Standalone side, we have conducted significant refactoring (#2794, #2820, #2809, #2774, #2807), including introducing a checkpoint mechanism and redesigning the storage mode for datastore meta files. These improvements result in a substantial reduction in the time required for the initial opening of a dataset while maintaining low memory consumption.

• Dataset: We have implemented reliability guarantees for the dataset loader session consumption mechanism(#2849, #2858, #2763, #2770, #2860, #2852, #2746), ensuring that with a large number of replicas, all dataset data will be consumed and no data loss will occur. Additionally, we have optimized the performance of dataset copying for large datasets(#2759, #2781) and supported mixed types in the list or tuple for dataset features(#2754). Furthermore, we have conducted performance optimizations for copying extensive datasets(#2759, #2781) and expanded our support to encompass mixed data types within lists or tuples when defining dataset features(#2754).

• Python SDK: It is now possible to create model evaluation tasks(#2839) for Standalone and Server/Cloud through the Python SDK. At the same time, we have redesigned the Evaluation SDK to enable the logging and scanning of data like images, videos, and audio to meet the need to record more complex evaluation metrics(#2808).

• UI: We have made many improvements in Web UI, including support for Job Templates (#2810, #2801), Advanced Search Bar (#2855), faster filtering of models/runtimes/datasets when creating Evaluations (#2736), and a streamlined Projects page (#2728).

• API Spec: We have redefined the API interface interaction between the Client and Server by leveraging the Pydantic library(#2639, #2753, #2734), thereby introducing an automated workflow for maintaining and updating the interface specifications. This initiative guarantees seamless compatibility across all interfaces.

💔 Breaking Changes

We must make breaking changes on the Standalone instance side for a better datastore design. Here are some operation advice:

• If you want to keep dataset data in the Standalone instance:
  • Step 1: Copy the dataset into the Server Side with the swcli dataset copy command.
  • Step 2: Run rm -rf ~/.starwhale/.datastore ~/.starwhale/self/dataset ~/.starwhale/self/job command.
  • Step 3: Upgrade Starwhale by pip install -U "starwhale>=0.6.0" command.
  • Step 4: Download the previously uploaded dataset.
• If you don't want any data in the Standalone instance:
  • Step 1: Run the rm -rf ~/.starwhale command.
  • Step 2: Upgrade Starwhale by pip install -U "starwhale>=0.6.0" command.

On the server side, full compatibility with legacy data has been achieved without requiring additional actions.

🧭 What's Changed

• refactor(client): add models and API for the requests (#2639) @jialeicui
• refactor(client): dataset related model (#2753) @jialeicui
• refactor(client): use pydantic model in runtime context (#2734) @jialeicui
• enhance(console): search bar with a large number of interaction revisions, including keyboard linkage/support hints selector (#2855) @waynelwz
• update(console): filter by only my project (#2728) @waynelwz
• feat(console): tree selector support quick retrieval by shared/current/latest (#2736) @waynelwz
• feat(console): support job as template (#2810) @waynelwz
• feat(controller): add job template api (#2801) @goldenxinxing
• feat(sdk): support evaluation log artifacts (#2808) @tianweidut
• feat(SDK): add job create sdk (#2839) @tianweidut
• enhance(client): optimize the predict count logging for greater accuracy (#2730) @tianweidut
• chore(client): tune dataset copy progress ui (#2759) @tianweidut
• enhance(dataset): optimize dataset copy performance to minimize redundant full table scans (#2781) @tianweidut
• feat(dataset): support mixed types in the list or tuple for dataset features (#2754) @tianweidut
• bug(SDK): fix loader consumption missing upload issue (#2849) @tianweidut
• chore(ci): enhance session consume test stability (#2858) @tianweidut
• ci(e2e): add huge-tasks example into e2e (#2763) @tianweidut
• enhance(dataset): tune processed keys for data loader (#2770) @tianweidut
• fix(controller): add max wait time for load request (#2860) @goldenxinxing
• fix(controller): cache unassigned range data to memory for optimize dataset consumption (#2852) @goldenxinxing
• fix(controller): reset unprocessed data when job is failed (#2746) @goldenxinxing
• chore(datastore): add tombstone and checkpoint to manifest (#2799) @jialeicui
• enhance(datastore): support tombstone in standalone (#2807) @jialeicui
• enhance(datastore): auto dump memory table and reduce memory usage when building dataset (#2794) @jialeicui
• feat(datastore): datastore in standalone support checkpoint (#2820) @jialeicui
• refactor(datastore): use flat structure to avoid too many subdirectories (#2809) @jialeicui
• optimize(datastore): speed up opening datastore (#2774) @jialeicui
• enhance(datastore): dump wal to the local disk before sending to the remote storage (#2782) @xuchuan
• chore(chart): add wal log cache support (#2800) @jialeicui
• enhance(datastore): allow dumping when updating tables (#2805) @xuchuan
• bug(datastore): parquet read/write issue with list/map of objects (#2825) @xuchuan
• chore(datastore): use revision instead of timestamp as record versions (#2804) @xuchuan
• bug(ci): fix no Makefile example for e2e (#2771) @tianweidut
• bug(example): fix nmt example typo (#2842) @tianweidut
• bug(SDK): fix list/tuple types check for converting sequence (#2837) @tianweidut
• bug(SDK): resolve the 'evaluate phase cannot read tables in standalone' issue (#2772) @tianweidut
• chore(SDK): fix job doc typo (#2731) @tianweidut
• chore(SDK): remove concurrency field in sdk/client side (#2769) @tianweidut
• chore(SDK): support memory resource for human (#2788) @tianweidut
• chore(ci): add huge-tasks model.yaml for e2e case (#2779) @tianweidut
• chore(ci): add open api model test in e2e test (#2751) @jialeicui
• chore(ci): disable e2e and codeql cron (#2826) @tianweidut
• chore(ci): fix e2e no disk space and TableWriter close issue (#2854) @tianweidut
• chore(ci): speedup e2e with disable huge-task in standalone (#2798) @tianweidut
• chore(ci): tune examples for e2e (#2780) @tianweidut
• chore(client): add double quotes to variables (#2806) @goldenxinxing
• chore(client): add retry condition and count limit for API requests (#2859) @jialeicui
• chore(client): config resource type for runtime uri (#2738) @tianweidut
• chore(client): fix mypy lint issue (#2844) @jialeicui
• chore(client): rm lock when dumping table (#2818) @jialeicui
• chore(client): support filter with the server api (#2749) @jialeicui
• chore(client): update TableWriter default limits of run exceptions for fast fail (#2831) @tianweidut
• chore(controller): add max thread property for server and tune default to 500 (#2846) @goldenxinxing
• chore(controller): modify k8s job name to a legal one (#2832) @anda-ren
• chore(controller): remove legacy job result query (#2817) @anda-ren
• chore(controller): unify artifact's creation conflict code (#2816) @goldenxinxing
• chore(dataset): fix dataset types default init value issue (#2757) @tianweidut
• chore(datastore): remove schema checking when query or scan (#2729) @jialeicui
• chore(e2e): rsync huge-tasks for e2e env prepare (#2787) @tianweidut
• chore(e2e): tune random runtime for examples (#2789) @tianweidut
• chore(helm): update pypi index url for minikube cn (#2775) @tianweidut
• chore(test): auto copy test use local dataset (#2851) @goldenxinxing
• chore(example): update huge-tasks example evaluate validation (#2863) @tianweidut
• doc: update GitHub readme (#2732) @tianweidut
• chore(e2e): remove upgrade test from e2e (#2722) @anda-ren
• enhance(client): add datastore log(evaluation, dataset) id type validation (#2776) @tianweidut
• enhance(client): add debug cli for data store (#2834) @jialeicui
• enhance(client): make building environment python version as the runtime default python version (#2819) @tianweidut
• enhance(dataset): expose exceptions from dataset copy threads (#2829) @tianweidut
• enhance(dataset): fix dataset flush block when it meets the unsupported types in the features (#2796) @tianweidut
• enhance(dataset): tune the length of dataset (#2803) @tianweidut
• example(LLM): support baichuan2 llm models (#2717) @tianweidut
• example(dataset): add coco ds sdk builder (#2745) @anda-ren
• example(dataset): add huge tasks example (#2740) @tianweidut
• feat(client): support dataset head rows for the prediction phase of model run command (#2792) @tianweidut
• feat(client): refine project uri in standalone (#2785) @goldenxinxing
• feat(client): support auto copy artifact when run or serve (#2823) @goldenxinxing
• feat(client): support custom Starwhale version in runtime yaml (#2711) @goldenxinxing
• feat(console): save column width/resize status to config (#2793) @waynelwz
• feat(controller): add eval upload api for file (#2695) @goldenxinxing
• feat(controller): add list api for run (#2815) @anda-ren
• feat(controller): add recentlyUsed flag for artifact vo (#2725) @goldenxinxing
• feat(controller): artifact's list API support search by name (#2735) @goldenxinxing
• feat(controller): controller support rolling update (#2760) @anda-ren
• feat(controller): suppo...

v0.5.12 🐋

🧭 What's Changed

• bug(ci): fix dockerfile name typo (#2715) @tianweidut
• bug(client): fix job info for server instances (#2662) @tianweidut
• bug(client): fix job list show error for server instance (#2658) @tianweidut
• bug(client): fix runtime quickstart unit test (#2633) @tianweidut
• bugfix(datastore): filter out empty records in query results (#2688) @xuchuan
• chore(ci): upgrade mypy related libs version (#2664) @tianweidut
• chore(ci): workaround for lfs quote issue to move dummy example lib to other repo (#2702) @tianweidut
• chore(client): update the default signed url expiration time (#2677) @tianweidut
• chore(console): tiptap editor default role caused the internal selector disabled, report preview error, etc (#2714) @waynelwz
• chore(controller): add stepSpec property for jobVo (#2631) @goldenxinxing
• chore(controller): enable compression for web server (#2699) @goldenxinxing
• chore(controller): refine docker related ut (#2645) @jialeicui
• chore(controller): show the exception of cmdline runner (#2697) @jialeicui
• chore(controller): sync step spec schema with client (#2704) @jialeicui
• chore(controller):tune dump period and add more logs (#2630) @goldenxinxing
• chore(helm): update controller k8s probe config (#2628) @tianweidut
• chore(workflow):add dataset builder to release action (#2638) @goldenxinxing
• chore: restore files to git from lfs (#2703) @tianweidut
• chroe: remove lfs files (#2709) @tianweidut
• example(LLM): add llm leaderboard analysis (#2705) @tianweidut
• example(LLM): add llm leaderboard examples (#2610) @tianweidut
• example(LLM): support llama2 chinese models (#2648) @tianweidut
• example(LLM): support tigerbot models (#2678) @tianweidut
• example(LLM): support xverse llm models (#2684) @tianweidut
• example(LLM): tune llm leaderboard for explanaion and readme (#2670) @tianweidut
• example: fix mnist memory limit typo (#2712) @tianweidut
• example: update mnist memory limit value (#2713) @tianweidut
• feat(SDK): PipelineHandler supports replicas and resources spec (#2701) @tianweidut
• feat(SDK): add Job SDK (#2653) @tianweidut
• feat(console): support eval table model multi select (#2687) @waynelwz
• feat(controller): add dataset/model/runtime base info in job (#2646) @goldenxinxing
• feat(controller): export full table result (#2707) @anda-ren
• feat: support model run handler with args (#2614) @anda-ren
• fix(chart): add default pypi configs in chart (#2627) @goldenxinxing
• fix(client): fix handler class error (#2629) @anda-ren
• fix(client): flush of dataset may stuck after putting failed (#2698) @jialeicui
• fix(client): prevent modification of built-in tags (#2671) @jialeicui
• fix(console): cache selected record in case of data missing when pagination (#2696) @waynelwz
• fix(console): async load content will be string, if direct save will cause double stringify occasionally (#2693) @waynelwz
• fix(console): e2e should have model of mnist with multi tag named mnist (#2637) @waynelwz
• fix(console): e2e text button to icon button (#2679) @waynelwz
• fix(console): error occurs when list model by project name (#2666) @jialeicui
• fix(console): make resource pool and step spec works in rerun (#2643) @jialeicui
• fix(console): new eval select logic/report header fixed etc (#2661) @waynelwz
• fix(console): support job form auto fill disable (#2657) @waynelwz
• fix(console): vscode error link (#2710) @waynelwz
• fix(controller): add chatset response header (#2683) @anda-ren
• fix(controller): fix built-in job shows on evaluation list (#2691) @anda-ren
• fix(controller): fix project count error (#2685) @anda-ren
• fix(controller): model path uri do not work (#2669) @jialeicui
• fix(controller): model related api support model id (#2665) @jialeicui
• fix(controller): prevent overflow when setting the default memory request (#2675) @jialeicui
• fix(controller): prevent readLine quit before read all log lines (#2652) @jialeicui
• fix(controller): report desc can't update to empty (#2692) @goldenxinxing
• fix(controller): resume can not work if delay deletion is configured (#2694) @jialeicui
• fix(controller): save to db failed on the stage of create job (#2673) @goldenxinxing
• fix(controller): tune report content datatype from text to mediumtext (#2635) @goldenxinxing
• fix(controller): use current user when create project (#2654) @jialeicui
• fix(controller): use lower case name in image build (#2636) @goldenxinxing
• hotfix(console): use state machine refactor multi job async auto fill state (#2676) @waynelwz
• optimize(console): panel widget sync with selected list, add editable status to panel widget, ui optimize etc (#2621) @waynelwz
• optimize(console): report interactive and support new chart config, title/sort/pagination/filterable etc (#2656) @waynelwz
• refactor(client): add job manifest model (#2659) @jialeicui
• refactor(controller): make controller independent on k8s thoroughly (#2689) @anda-ren
• refactor(controller): make ds build a job (#2626) @anda-ren
• refactor(controller): make online evaluation be a job (#2672) @anda-ren
• refactor(controller): make runtime dockerizing be a job (#2651) @anda-ren
• refactor(deploy): exclude minikube from docker compose (#2690) @anda-ren
• update(console): config sync、reset status between chart and config modal , optimize ui (#2650) @waynelwz
• update(console): decrease page size, min width for ploty chart (#2700) @waynelwz
• update(console): monaco editor disable remote load from cdn (#2663) @waynelwz
• update(console): new pagination ui & fix export without schema & grid table with tuple schema (#2706) @waynelwz
• update(console): optimize report & panel interactive (#2632) @waynelwz
• update(console): panel report optimize rerender/chart edit save/sort etc (#2642) @waynelwz
• update(console): report readonly mode remove edit actions (#2708) @waynelwz
• update(console): show share button when shared / all list page actions to icons (#2668) @waynelwz
• update(console): support local sort / delete from list / job overview etc (#2686) @waynelwz

⚙️ Who Contributes

@anda-ren, @goldenxinxing, @jialeicui, @tianweidut, @waynelwz and @xuchuan

v0.5.11 🐋

🧭 What's Changed

• chore(client): runtime build does not dump pip and conda configs by default (#2601) @tianweidut
• chore(console): refine tag related pages (#2599) @jialeicui
• chore(controller): add editorconfig for server (#2616) @jialeicui
• chore(controller): add flatten properties for artifact vo (#2613) @goldenxinxing
• chore(controller): refine resource info vo (#2598) @jialeicui
• chore(e2e): add white list for e2e clean script (#2622) @goldenxinxing
• enhance(client): enhance dataset build from huggingface for subsets (#2608) @tianweidut
• enhance(client): refactor model/dataset/runtime list for server and cloud instance (#2625) @tianweidut
• example(LLM): build llama2 other base models (#2533) @tianweidut
• feat(client): support build dataset from csv (#2602) @tianweidut
• feat(client): support dataset build from jsonl files (#2606) @tianweidut
• feat(console): report list card (#2600) @jialeicui
• feat(console): support wysiwyg report editor (#2618) @waynelwz
• feat(controller): add table scan export (#2612) @anda-ren
• feat(controller):add report api (#2605) @goldenxinxing
• fix(console): disable task succeed status for delay case (#2603) @waynelwz
• fix(console): make runtime in rerun works (#2607) @jialeicui
• fix(console): tag edit role access control (#2609) @jialeicui
• fix(console): update model case (#2597) @waynelwz
• fix(controller): report revert error (#2615) @goldenxinxing
• hotfix(console): ignore extensions (#2624) @waynelwz
• refactor(controller): refine the abstract layer between job and k8s (#2604) @anda-ren

⚙️ Who Contributes

@anda-ren, @goldenxinxing, @jialeicui, @tianweidut and @waynelwz

v0.5.10 🐋

🧭 What's Changed

• bug(charts): fix minio svc name typo (#2595) @tianweidut
• bug(ci): fix server api test typo for tag api (#2594) @tianweidut
• bug(ci): fix tag check in the api test (#2596) @tianweidut
• chore(client): Support Resourse.refine for the user's tags in the standalone instance (#2586) @tianweidut
• chore(client): update conda log verbose mapping to reduce the useless output (#2582) @tianweidut
• chore(console): link to the latest version in model list (#2578) @jialeicui
• chore(console): show web handler links in task list view (#2576) @jialeicui
• chore(controller): add more detail info for job api (#2571) @goldenxinxing
• chore(controller): support exposed link in task list api (#2574) @jialeicui
• chore(controller): validate tag name and add api for getting version id by tag (#2581) @jialeicui
• chore(script): record sync version to file (#2561) @goldenxinxing
• enhance(console): support custom tag (#2588) @jialeicui
• enhance(controller): support baidu oss offical sdk (#2519) @jialeicui
• enhance(controller): support custom tags for the resources (#2569) @jialeicui
• example(LLM): disable save checkpoint for llama finetune (#2585) @tianweidut
• example(LLM): fix llama finetune build model modules typo (#2587) @tianweidut
• feat(client): support copy with tags for model/runtime/dataset (#2589) @tianweidut
• feat(client): support --tag argument for model/runtime/dataset build (#2580) @tianweidut
• feat(client): support swcli model/dataset/runtime tag for the server instance (#2590) @tianweidut
• feat(sdk): add login and logout sdk (#2579) @tianweidut
• feat: add support for setting up a server instance in the manner of docker compose (#2584) @anda-ren
• fix(controller): add constraint(UK) for resource entity (#2568) @goldenxinxing
• fix(controller): fix the missing function call (#2593) @jialeicui
• update(console): control render rate avoid user interactive blocking, cached component (#2583) @waynelwz
• update(console): refactor job form combint extend components, support rerun & online eval (#2572) @waynelwz
• update(console): remove modify check when rerun (#2577) @waynelwz

⚙️ Who Contributes

@anda-ren, @goldenxinxing, @jialeicui, @tianweidut and @waynelwz

v0.5.9 🐋

🧭 What's Changed

• bug(client): fix base/cuda image typo (#2575) @tianweidut
• bug(client): fix commands help typo (#2567) @tianweidut
• chore(SDK): update sdk help docs and fix fine_tune needs typo (#2570) @tianweidut
• chore(chart):update chart desc and server image tag in e2e (#2554) @goldenxinxing
• chore(docker): add Dockerfile for dataset build image (#2562) @goldenxinxing
• feat(controller): add header for exported csv (#2564) @anda-ren
• fix(controller): do not validate the resource with name in filter (#2563) @jialeicui
• fix(controller): online log with the correct pod (#2565) @jialeicui
• update(console): replace inject html (#2560) @waynelwz
• update(console): widget full reload to module reload in vite hmr 、 baidu analytics (#2559) @waynelwz

⚙️ Who Contributes

@anda-ren, @goldenxinxing, @jialeicui, @tianweidut and @waynelwz