RAGFlow-0.13.0

Summary

• Added team support
• Upgrade UI for Agent
• Added HTTP API for knowledge base, document parsing, dialogs, etc. Python SDK is also provided, which can be installed via pip install ragflow-sdk==0.13.0.
• Added smart chunking for Markdown
• Added context retrieval through entity extraction from LLM
• Added Invoke operator for Agent
• Added support for DIFY external knowledge base API
• Added integration for ChatGPT-on-WeChat
• Added support for GLM4-9B, Yi-Lightning

What's Changed

• added tests.yml by @yuzhichang in #2685
• fix tests.yml by @yuzhichang in #2688
• refine readme by @KevinHuSh in #2689
• format mind_map_extractor code by @yqkcn in #2686
• Simplify the usage of dict by @yqkcn in #2681
• remove unused import by @yqkcn in #2679
• Refactoring entity_resolution by @guoyuhao2330 in #2692
• fix error in exception by @KevinHuSh in #2694
• Updated Dockerfile by @yuzhichang in #2695
• fix: Fixed the issue where no error message was displayed when uploading a file that was too large #2258 by @cike8899 in #2697
• Updated Dockefile to use cache by @yuzhichang in #2703
• Fix Dockerfile for arm64 by @yuzhichang in #2705
• Updated Build Docker Image by @writinwaters in #2706
• Fix Dockerfile.slim by @yuzhichang in #2708
• Fix poetry cache by @yuzhichang in #2709
• docs: Migrate README to Compose V2 syntax by @anud18 in #2711
• build multi-arch image by @yuzhichang in #2710
• Fix/bedrock issues by @Omarlsg98 in #2718
• remove redeclared get_json_result func by @yqkcn in #2675
• fix re.escape problem by @KevinHuSh in #2735
• chore: update operators.py by @eltociear in #2724
• support sequence2txt and tts model in Xinference by @JobSmithManipulation in #2696
• Add query parts to lamp by @KevinHuSh in #2736
• feat: Add component TuShare #1739 by @cike8899 in #2745
• Fix LIGHTEN. Close #2723 by @yuzhichang in #2744
• expand rerank range by @KevinHuSh in #2746
• Modified download_deps.py by @yuzhichang in #2747
• Update azure_sas_conn.py - fixing container_url typo by @jstride in #2740
• solve knowledgegraph issue when calling gemini model by @JobSmithManipulation in #2738
• move import positions by @KevinHuSh in #2753
• feat: The same query conditions on the search page should not request the interface every time the mind map drawer is opened. #2759 by @cike8899 in #2760
• Initial draft of Python APIs by @writinwaters in #2719
• Fix component exesql by @guoyuhao2330 in #2754
• fix xinference add rerank model bug by @wangguo1230 in #2758
• Fix docx images by @guoyuhao2330 in #2756
• Docs:fix user guide links by @Howe829 in #2761
• feat: Add a note node to the agent canvas #2767 by @cike8899 in #2768
• HTTP API documents, part1 by @JinHai-CN in #2713
• Use slim image as the default by @JinHai-CN in #2772
• Fix README description by @JinHai-CN in #2773
• Fix README and some comments by @JinHai-CN in #2774
• fix: Large document thumbnail display failed by @chongchuanbing in #2763
• fix bedrock issue by @KevinHuSh in #2776
• fix: torch dependency start error by @chongchuanbing in #2777
• Fix error commands by @JinHai-CN in #2778
• trival by @KevinHuSh in #2779
• Refactor README on different docker version. by @JinHai-CN in #2775
• refine agent by @KevinHuSh in #2787
• Updated instructions on downloading RAGFlow Slim and RAGFlow all-in-one. by @writinwaters in #2785
• fix: Change document status with @tanstack/react-query #13306 by @cike8899 in #2788
• Refactor Dataset API by @Feiue in #2783
• feat: Arrange the form of generate, categorize, switch, retrieval operators vertically #1739 by @cike8899 in #2800
• support api-version and change default-model in adding azure-openai and openai by @JobSmithManipulation in #2799
• fix: document preview by @chongchuanbing in #2795
• reduce rerank batch size by @KevinHuSh in #2801
• Fixed broken links. by @writinwaters in #2805
• support chat model in huggingface by @JobSmithManipulation in #2802
• fix LIGHTEN issue by @KevinHuSh in #2806
• trival by @KevinHuSh in #2808
• fix chat and thumbnail bug by @chongchuanbing in #2803
• Refactor Chat API by @Feiue in #2804
• refine components retrieval and rewrite by @KevinHuSh in #2818
• Refactor API for document and session by @Feiue in #2819
• Updated dataset APIs by @writinwaters in #2820
• refine token similarity by @KevinHuSh in #2824
• feat: Add message_history_window_size to CategorizeForm #1739 by @cike8899 in #2826
• feat: Fix translation issue of message_history_window_size #1739 by @cike8899 in #2828
• Refactor Document API by @Feiue in #2833
• Updated chat APIs by @writinwaters in #2831
• let presentation do raptor by @KevinHuSh in #2838
• feat: Generate operator names in an auto-incremental manner #1739 by @cike8899 in #2844
• refine API token application by @KevinHuSh in #2847
• refactor api util by @KevinHuSh in #2849
• add rm token by @KevinHuSh in #2850
• Feat:Compatible with Dify's External Knowledge API by @Howe829 in #2848
• Enhance email validation: Allow top-level domains with 5 letters by @mrnicegyu11 in #2856
• refactor by @KevinHuSh in #2858
• Fix keys of Xinference deployed models, especially has the same model name with public hosted models. by @0000sir in #2832
• feat: API access key management #2846 by @cike8899 in #2865
• add API for tenant function by @KevinHuSh in #2866
• Refactor Chunk API by @Feiue in #2855
• Updated session APIs by @writinwaters in #2868
• update dashscope by @KevinHuSh in #2871
• feat: Configure the root directory alias #1739 by @cike8899 in #2875
• feat: Load markdown file with "asset/source" #1739 by @cike8899 in #2880
• Fixed docker build by @yuzhichang in #2881
• Draft: Updated file management-related APIs by @writinwaters in #2882
• Miscellaneous minor updates by @writinwaters in #2885
• feat: Bind data to TenantTable #2846 by @cike8899 in #2883
• add team tag to kb by @KevinHuSh in #2890
• Delete useless files by @Feiue in #2889
• feat: Use Badge.Ribbon to distinguish the teams to which the knowledge base belongs #2846 by @cike8899 in #2891
• add owner check for team work by @KevinHuSh in #2892
• fix: Fixed the issue of error reporting when uploading files in the chat box #2897 by @cike8899 in #2898
• DRAFT: Updated chunk APIs by @writinwaters in #2901
• DRAFT: Miscellaneous proofedits on Python APIs by @writinwaters in #2903
• add nginx path for sdk handlers (#2899) by @kaleocheng in #2900
• DRAFT: Miscellaneous updates to HTTP API. Tried to finish off Python API ref… by @writinwaters in #2909
• Resolves #2905 openai compatible model provider add llama.cpp rerank support by @ziyu4huang in #2906
• add zhipu glm-4-9b by @KevinHuSh in #2912
• Add agent component for web crawler by @wingjson in #2878
• Feature/feat1017 by @chongchuanbing in #2872
• feat: Replace crawler icon #2915 by @cike8899 in #2916
• Fix some issues in API by @Feiue in #2902
• fix thumbnail issue by @KevinHuSh in #2917
• add package crawl4ai by @KevinHuSh in #2918
• bigger resolution for OCR by @KevinHuSh in #2919
• DRAFT: Miscellaneous updates to HTTP API reference by @writinwaters in #2923
• Fix thumbnail_img NoneType error by @wingjson in #2941
• fix sequence2txt error and usage total token issue by @KevinHuSh in #2961
• [Bug]: When use OpenAI chat model , raise ERROR: 'CompletionUsage' object has no attribute 'get' #2948 by @yqwang96 in #2949
• add auto keywords and auto-question by @KevinHuSh in #2965
• add component invoke by @KevinHuSh in #2967
• make titles in markdown not be splited with following content by @KevinHuSh in #2971
• [Bug]: unnecessary auto-increment calculations in the tokens statistics of the chat model by @yqwang96 in #2969
• DRAFT: Updated python and http api references by @writinwaters in #2973
• fix SILICONFLOW rerank error by @KevinHuSh in #2980
• fix generate string join issue by @KevinHuSh in #2983
• minor by @writinwaters in #2984
• Fix some issues in API by @Feiue in #2982
• feat: Added auto_keywords and auto_questions fields to the parsing configuration page #2687 by @cike8899 in #2987
• allowing docker container to access service on host by @benquike in #2895
• refactor auto keywords and auto question by @KevinHuSh in #2990
• Updated HTTP API Reference (document, chat assistant, session, chat) by @writinwaters in #2994
• Test Cases by @Feiue in #2993
• Remove defaults to 'None' by @JinHai-CN in #2996
• Integrating RAGFlow API as a Plugin into ChatGPT-on-WeChat by @timebankspace in #2988
• feat: Limit the maximum value of auto keywords to 30 #2687 by @cike8899 in #2991
• minor by @writinwaters in #2998
• Miscellaneous updates to HTTP and PYthon APIs by @writinwaters in #3000
• Miscellaneous updates to HTTP API Reference by @writinwaters in #3005
• Fix some issues in API and test by @Feiue in #3001
• Fix issues in API by @Feiue in #3008
• make gemini robust by @KevinHuSh in #3012
• Final touches to HTTP and Python API references by @writinwaters in #3019
• Fix non-null violation when inviting people to team by @b246wang in #3015
• Minor editorial updates to the HTTP API reference by @writinwaters in #3027
• add model moonshot-v1-auto by @KevinHuSh in #3051
• fix baidufanyi param error by @KevinHuSh in #3053
• Bump werkzeug from 3.0.3 to 3.0.6 by @dependabot in #3026
• Turn resource to plural form by @Feiue in #3059
• feat: Add hint for operators, round to square, input variable, readable operator ID. #3056 by @cike8899 in #3057
• Turn resource to plural form by @Feiue in #3061
• agent template upgrade by @KevinHuSh in #3060
• remove file size check for sdk API by @KevinHuSh in #3066
• feat: Add component Invoke #2908 by @cike8899 in #3067
• refactor error message of qwen by @KevinHuSh in #3074
• fix local variable ans by @KevinHuSh in #3077
• Update 'api key' by @JinHai-CN in #3078
• feat: If the model supplier is not set, click the OK button to jump directly to the page for setting the model supplier. #3068 by @cike8899 in #3069
• fix: modify the response of metadata in Dify retrieval api by @k0286 in #3076
• search between multiple indiices for team function by @KevinHuSh in #3079
• feat: Add InvokeNode #1908 by @cike8899 in #3081
• let 'Generate' take user's input as parameter by @KevinHuSh in #3086
• feat: Allow the component id drop-down box to select the answer operator #3085 by @cike8899 in #3087
• feat: The order of the category operator form is messed up after refreshing the page #3088 by @cike8899 in #3089
• Updated HTTP API reference and Python API reference based on test results by @writinwaters in #3090
• Update README by @JinHai-CN in #3092
• Updated parser_config description by @writinwaters in #3104
• Fix bugs in API by @Feiue in #3103
• Differentiated API key names to avoid confusion. by @writinwaters in #3107
• Added chunk methods by @writinwaters in #3110
• docs updates for 0.13 release by @KevinHuSh in #3108
• Publish RAGFlow's HTTP and Python API references by @writinwaters in #3116
• Tried to fix a link issue by @writinwaters in #3117
• Manage ragflow-sdk with poetry by @yuzhichang in #3115
• add yi-lightning by @KevinHuSh in #3119
• Fixed a docusaurus display issue by @writinwaters in #3120
• feat: Delete http_api_reference.md from api folder #1102 by @cike8899 in #3121
• Fixed a docusaurus display issue. by @writinwaters in #3124
• fix: remove useless test code by @zhyg in #3122
• Fixed a Docusaurus display issue. by @writinwaters in #3125
• Add test for CI by @Feiue in #3114
• Updated list_documents() by @writinwaters in #3126
• Moved the Upgrade Manuel out of FAQ by @writinwaters in #3131
• Reworded descriptions for development versions and latest version by @writinwaters in #3132
• minor by @writinwaters in #3137

New Contributors

• @anud18 made their first contribution in #2711
• @Omarlsg98 made their first contribution in #2718
• @jstride made their first contribution in #2740
• @wangguo1230 made their first contribution in #2758
• @Howe829 made their first contribution in #2761
• @chongchuanbing made their first contribution in #2763
• @mrnicegyu11 made their first contribution in #2856
• @0000sir made their first contribution in #2832
• @kaleocheng made their first contribution in #2900
• @ziyu4huang made their first contribution in #2906
• @yqwang96 made their first contribution in #2949
• @benquike made their first contribution in #2895
• @timebankspace made their first contribution in #2988
• @b246wang made their first contribution in #3015
• @k0286 made their first contribution in #3076
• @zhyg made their first contribution in #3122

Full Changelog: v0.12.0...v0.13.0