Releases · pingcap/autoflow

03 Jan 09:33

Mini256

0.4.0

b12447b

0.4.0 Latest

Latest

Highlights

Support Evaluation (beta) tool to evaluate the performance and reliability of the Chat Engine’s outputs (by @Icemap @634750802 ).
- Current support key metrics:
  - Factual Correctness
  - Semantic Similarity
Support new LLM providers
- Gitee AI (by @lc-soft)
- Test new OpenAI-like providers (by @jrj5423)
  - vLLM
  - Xinference
Support new embedding model providers
- Gitee AI (by @lc-soft)
- Amazon Bedrock (by @jrj5423 @Mini256)

Improvements

Limit the upload file size via max_upload_file_size (10MB by default) parameter on site setting (by @sszgwdk @634750802 )
Support download the reference file in the chat page (by @Rutheniumlmw @634750802 )

New Contributors

@sszgwdk made their first contribution in #497
@jrj5423 made their first contribution in #522
@Rutheniumlmw made their first contribution in #538
@lc-soft made their first contribution in #470
@nicognaW made their first contribution in #496

If you are deploying Autoflow using docker, please follow the Upgrade guide to upgrade your Autoflow.

Full Changelog: v0.3.0...0.4.0

Contributors

lc-soft, Mini256, and 6 other contributors

Assets 2

31 Dec 07:24

Mini256

0.4.0rc1

85872a9

0.4.0rc1 Pre-release

Pre-release

What's Changed

e2e: fix bootstrap by @634750802 in #438
fix: structured prediction is only for openai api by @Icemap in #487
ui(frontend): add a cool homepage background made by magicui.design by @634750802 in #464
ui(frontend): refine progress chart label by @634750802 in #489
chore(frontend): upgrade to nextjs 15 + react 19 by @634750802 in #488
e2e: fix bootstrap by @634750802 in #491
fix(frontend): refresh after navigating by @634750802 in #492
feat: evaluate api by @Icemap in #463
feat(frontend): evaluation management pages by @634750802 in #495
Enable passing custom domain username via cli by @chethanuk in #420
add verification for stackvm by @IANTHEREAL in #503
feat: add site setting upload_max_body_size by @sszgwdk in #497
Enable passing custom admin username and password, enable resetting … by @sszgwdk in #499
support response format by @IANTHEREAL in #504
fix: staging errors by @Icemap in #498
feat: update evaluation api group by @Icemap in #515
feat(frontend): refine evaluation management pages by @634750802 in #513
Extend graph search capabilities by @IANTHEREAL in #517
refine(frontend): partial replace react-hook-form with @tanstack/form by @634750802 in #519
Improve graph basic api by @IANTHEREAL in #521
docs: add examples of vLLM and Xinference models by @jrj5423 in #522
refactor(frontend): replace react-hook-form with @tanstack/form by @634750802 in #524
Refine the goal background generation and api by @IANTHEREAL in #530
e2e: add new chat engine tests by @634750802 in #529
fix(frontend): only show widget on homepage by @634750802 in #531
fix(frontend): fix widget display conditions by @634750802 in #533
e2e: add knowledge base test cases by @634750802 in #536
doc: evaluation by @Icemap in #539
docs: add "How to add an API?" to CONTRIBUTING.md by @Rutheniumlmw in #538
Upgrade claude sonnet models by @chethanuk in #534
Fix embedding model bug and support arm platform by @IANTHEREAL in #543
feat: change the visibility of evaluation entities by @Icemap in #540
fix bug of merging entities by @IANTHEREAL in #546
feat: add Gitee AI provider by @lc-soft in #470
Add current to text refine question prompt by @nicognaW in #496
e2e: add evaluation dataset test cases by @634750802 in #544
fix chunks model for knowledge graph by @IANTHEREAL in #550
docs: fix dev docker compose filepath on CONTRIBUTING.md by @lc-soft in #552
docs: update the LLM / Embedding models for Gitee AI by @lc-soft in #551
feat: support download uploaded documents by @Rutheniumlmw in #532
feat(frontend): support download uploaded files by @634750802 in #553
feat: support bedrock embedding model provider by @jrj5423 in #511