version to 0.6.1 (#3253 )

feat: remove unregistered-llm-in-debug (#3251 )
make sure validation flow works for all model providers in bedrock (#3250 )
2024-04-09 21:21:09 +08:00 · 2024-04-09 20:49:52 +08:00 · 2024-04-09 20:42:18 +08:00 · 2024-04-09 19:28:22 +08:00 · 2024-04-09 19:25:08 +08:00 · 2024-04-09 18:33:47 +08:00
769 changed files with 58139 additions and 183243 deletions
@@ -8,6 +8,8 @@ body:
      label: Self Checks
      description: "To make sure we get to you in time, please check the following :)"
      options:
+        - label: This is only for bug report, if you would like to ask a quesion, please head to [Discussions](https://github.com/langgenius/dify/discussions/categories/general).
+          required: true
        - label: I have searched for existing issues [search for existing issues](https://github.com/langgenius/dify/issues), including closed ones.
          required: true
        - label: I confirm that I am using English to submit this report (我已阅读并同意 [Language Policy](https://github.com/langgenius/dify/issues/1542)).
@@ -1,8 +1,5 @@
 blank_issues_enabled: false
 contact_links:
-  - name: "\U0001F4DA Dify user documentation"
-    url: https://docs.dify.ai/getting-started/readme
-    about: Documentation for users of Dify
-  - name: "\U0001F4DA Dify dev documentation"
-    url: https://docs.dify.ai/getting-started/install-self-hosted
-    about: Documentation for people interested in developing and contributing for Dify
+  - name: "\U0001F4E7 Discussions"
+    url: https://github.com/langgenius/dify/discussions/categories/general
+    about: General discussions and request help from the community
@@ -1,22 +0,0 @@
-name: "🤝 Help Wanted"
-description: "Request help from the community [please use English :）]"
-labels:
-  - help-wanted
-body:
-  - type: checkboxes
-    attributes:
-      label: Self Checks
-      description: "To make sure we get to you in time, please check the following :)"
-      options:
-        - label: I have searched for existing issues [search for existing issues](https://github.com/langgenius/dify/issues), including closed ones.
-          required: true
-        - label: I confirm that I am using English to submit this report (我已阅读并同意 [Language Policy](https://github.com/langgenius/dify/issues/1542)).
-          required: true
-        - label: "Pleas do not modify this template :) and fill in all the required fields."
-          required: true
-  - type: textarea
-    attributes:
-      label: Provide a description of the help you need
-      placeholder: Briefly describe what you need help with.
-    validations:
-      required: true
@@ -12,7 +12,7 @@ Please delete options that are not relevant.
 - [ ] New feature (non-breaking change which adds functionality)
 - [ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)
 - [ ] This change requires a documentation update, included: [Dify Document](https://github.com/langgenius/dify-docs)
- [ ] Improvement，including but not limited to code refactoring, performance optimization, and UI/UX improvement
+- [ ] Improvement, including but not limited to code refactoring, performance optimization, and UI/UX improvement
 - [ ] Dependency upgrade

 # How Has This Been Tested?
@@ -26,6 +26,7 @@ jobs:
      HUGGINGFACE_TEXT2TEXT_GEN_ENDPOINT_URL: b
      HUGGINGFACE_EMBEDDINGS_ENDPOINT_URL: c
      MOCK_SWITCH: true
+      CODE_MAX_STRING_LENGTH: 80000

    steps:
      - name: Checkout code
@@ -41,5 +42,11 @@ jobs:
      - name: Install dependencies
        run: pip install -r ./api/requirements.txt

-      - name: Run pytest
+      - name: Run ModelRuntime
        run: pytest api/tests/integration_tests/model_runtime/anthropic api/tests/integration_tests/model_runtime/azure_openai api/tests/integration_tests/model_runtime/openai api/tests/integration_tests/model_runtime/chatglm api/tests/integration_tests/model_runtime/google api/tests/integration_tests/model_runtime/xinference api/tests/integration_tests/model_runtime/huggingface_hub/test_llm.py
+
+      - name: Run Tool
+        run: pytest api/tests/integration_tests/tools/test_all_provider.py
+
+      - name: Run Workflow
+        run: pytest api/tests/integration_tests/workflow
@@ -1,26 +0,0 @@
-name: Run Tool Pytest
-
-on:
-  pull_request:
-    branches:
-      - main
-
-jobs:
-  test:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout code
-        uses: actions/checkout@v4
-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.10'
-          cache: 'pip'
-          cache-dependency-path: ./api/requirements.txt
-
-      - name: Install dependencies
-        run: pip install -r ./api/requirements.txt
-
-      - name: Run pytest
-        run: pytest ./api/tests/integration_tests/tools/test_all_provider.py
@@ -1,31 +0,0 @@
-name: Run Pytest
-
-on:
-  pull_request:
-    branches:
-      - main
-      - deploy/dev
-
-jobs:
-  test:
-    runs-on: ubuntu-latest
-
-    env:
-      MOCK_SWITCH: true
-
-    steps:
-      - name: Checkout code
-        uses: actions/checkout@v4
-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.10'
-          cache: 'pip'
-          cache-dependency-path: ./api/requirements.txt
-
-      - name: Install dependencies
-        run: pip install -r ./api/requirements.txt
-
-      - name: Run pytest
-        run: pytest api/tests/integration_tests/workflow
@@ -5,7 +5,6 @@ on:
    branches:
      - "main"
      - "deploy/dev"
-      - "feat/workflow"
  release:
    types: [published]

@@ -47,7 +46,7 @@ jobs:
        with:
          images: ${{ env[matrix.image_name_env] }}
          tags: |
-            type=raw,value=latest,enable=${{ github.ref == 'refs/heads/main' && startsWith(github.ref, 'refs/tags/') }}
+            type=raw,value=latest,enable=${{ startsWith(github.ref, 'refs/tags/') }}
            type=ref,event=branch
            type=sha,enable=true,priority=100,prefix=,suffix=,format=long
            type=raw,value=${{ github.ref_name }},enable=${{ startsWith(github.ref, 'refs/tags/') }}
@@ -36,7 +36,7 @@ In terms of licensing, please take a minute to read our short [License and Contr
  | Feature Type                                                 | Priority        |
  | ------------------------------------------------------------ | --------------- |
  | High-Priority Features as being labeled by a team member     | High Priority   |
-  | Popular feature requests from our [community feedback board](https://feedback.dify.ai/) | Medium Priority |
+  | Popular feature requests from our [community feedback board](https://github.com/langgenius/dify/discussions/categories/feedbacks) | Medium Priority |
  | Non-core features and minor enhancements                     | Low Priority    |
  | Valuable but not immediate                                   | Future-Feature  |

@@ -155,4 +155,4 @@ And that's it! Once your PR is merged, you will be featured as a contributor in

 ## Getting Help

-If you ever get stuck or got a burning question while contributing, simply shoot your queries our way via the related GitHub issue, or hop onto our [Discord](https://discord.gg/AhzKf7dNgk) for a quick chat. 
+If you ever get stuck or got a burning question while contributing, simply shoot your queries our way via the related GitHub issue, or hop onto our [Discord](https://discord.gg/8Tpq4AcN9c) for a quick chat. 
@@ -34,7 +34,7 @@
  | Feature Type                                                 | Priority        |
  | ------------------------------------------------------------ | --------------- |
  | High-Priority Features as being labeled by a team member     | High Priority   |
-  | Popular feature requests from our [community feedback board](https://feedback.dify.ai/) | Medium Priority |
+  | Popular feature requests from our [community feedback board](https://github.com/langgenius/dify/discussions/categories/feedbacks) | Medium Priority |
  | Non-core features and minor enhancements                     | Low Priority    |
  | Valuable but not immediate                                   | Future-Feature  |

@@ -152,4 +152,4 @@ Dify的后端使用Python编写，使用[Flask](https://flask.palletsprojects.co

 ## 获取帮助

-如果你在贡献过程中遇到困难或者有任何问题，可以通过相关的 GitHub 问题提出你的疑问，或者加入我们的 [Discord](https://discord.gg/AhzKf7dNgk) 进行快速交流。
+如果你在贡献过程中遇到困难或者有任何问题，可以通过相关的 GitHub 问题提出你的疑问，或者加入我们的 [Discord](https://discord.gg/8Tpq4AcN9c) 进行快速交流。
@@ -1,4 +1,4 @@
-[![](./images/describe.png)](https://dify.ai)
+[![](./images/GitHub_README_cover.png)](https://dify.ai)
 <p align="center">
  <a href="./README.md">English</a> |
  <a href="./README_CN.md">简体中文</a> |
@@ -27,18 +27,37 @@
  </a>
 </p>

-**Dify** is an LLM application development platform that has helped built over **100,000** applications. It integrates BaaS and LLMOps, covering the essential tech stack for building generative AI-native applications, including a built-in RAG engine. Dify allows you to **deploy your own version of Assistants API and GPTs, based on any LLMs.**
+**Dify** is an open-source LLM app development platform. Dify's intuitive interface combines a RAG pipeline, AI workflow orchestration, agent capabilities, model management, observability features and more, letting you quickly go from prototype to production.

-![](./images/demo.png)
+https://github.com/langgenius/dify/assets/13230914/979e7a68-f067-4bbc-b38e-2deb2cc2bbb5


+## Using Dify Cloud

-## Using our Cloud Services
+You can try out [Dify Cloud](https://dify.ai) now. It provides all the capabilities of the self-deployed version, and includes 200 free GPT-4 calls.

-You can try out [Dify.AI Cloud](https://dify.ai) now. It provides all the capabilities of the self-deployed version, and includes 200 free requests to OpenAI GPT-3.5.
+## Dify for Enterprise / Organizations
+
+[Schedule a meeting with us](#Direct-Meetings) or [send us an email](mailto:[email protected]?subject=[GitHub]Business%20License%20Inquiry) to discuss enterprise needs. 
+
+For startups and small businesses using AWS, check out [Dify Premium on AWS Marketplace](https://aws.amazon.com/marketplace/pp/prodview-t22mebxzwjhu6) and deploy it to your own AWS VPC with one-click. It's an affordable AMI offering with the option to create apps with custom logo and branding.
+
+## Features
+
+![](./images/models.png)
+
+**1. Workflow**: Create and test complex AI workflows on a visual canvas, with pre-built nodes taking advantage of the power of all the following features and beyond.
+
+**2. Extensive LLM support**: Seamless integration with hundreds of proprietary / open-source LLMs and dozens of inference providers, including GPT, Mistral, Llama2, and OpenAI API-compatible models. A full list of supported model providers is kept [here](https://docs.dify.ai/getting-started/readme/model-providers).
+
+**3. Prompt IDE**: Visual orchestration of applications and services based on any LLMs. Easily share with your team.
+
+**4. RAG Engine**: Includes various RAG capabilities based on full-text indexing or vector database embeddings, allowing direct upload of PDFs, TXTs, and other text formats.
+
+**5. AI Agent**: Based on Function Calling and ReAct, the Agent inference framework allows users to customize tools, what you see is what you get. Dify provides more than a dozen built-in tools for AI agents, such as Google Search, DELL·E, Stable Diffusion, WolframAlpha, etc.
+
+**6. LLMOps**: Monitor and analyze application logs and performance, continuously improving Prompts, datasets, or models based on production data.

-### Looking to purchase via AWS?
-Check out [Dify Premium on AWS](https://aws.amazon.com/marketplace/pp/prodview-t22mebxzwjhu6) and deploy it to your own AWS VPC with one-click. 

 ## Dify vs. LangChain vs. Assistants API

@@ -52,28 +71,10 @@ Check out [Dify Premium on AWS](https://aws.amazon.com/marketplace/pp/prodview-t
 | **Local Deployment** | Supported | Not Supported | Not Applicable |


-
-## Features
-
-![](./images/models.png)
-
-**1. LLM Support**: Integration with OpenAI's GPT family of models, or the open-source Llama2 family models. In fact, Dify supports mainstream commercial models and open-source models (locally deployed or based on MaaS).
-
-**2. Prompt IDE**: Visual orchestration of applications and services based on LLMs with your team.
-
-**3. RAG Engine**: Includes various RAG capabilities based on full-text indexing or vector database embeddings, allowing direct upload of PDFs, TXTs, and other text formats.
-
-**4. AI Agent**: Based on Function Calling and ReAct, the Agent inference framework allows users to customize tools, what you see is what you get. Dify provides more than a dozen built-in tool calling capabilities, such as Google Search, DELL·E, Stable Diffusion, WolframAlpha, etc.
-
-
-**5. Continuous Operations**: Monitor and analyze application logs and performance, continuously improving Prompts, datasets, or models using production data.
-
 ## Before You Start

 **Star us on GitHub, and be instantly notified for new releases!**
-
 ![star-us](https://github.com/langgenius/dify/assets/100913391/95f37259-7370-4456-a9f0-0bc01ef8642f)
-
 - [Website](https://dify.ai)
 - [Docs](https://docs.dify.ai)
 - [Deployment Docs](https://docs.dify.ai/getting-started/install-self-hosted)
@@ -100,10 +101,12 @@ docker compose up -d

 After running, you can access the Dify dashboard in your browser at [http://localhost/install](http://localhost/install) and start the initialization installation process.

-### Helm Chart
+#### Deploy with Helm Chart

-Big thanks to @BorisPolonsky for providing us with a [Helm Chart](https://helm.sh/) version, which allows Dify to be deployed on Kubernetes.
-You can go to https://github.com/BorisPolonsky/dify-helm for deployment information.
+[Helm Chart](https://helm.sh/) version, which allows Dify to be deployed on Kubernetes.
+
+- [Helm Chart by @LeoQuote](https://github.com/douban/charts/tree/master/charts/dify)
+- [Helm Chart by @BorisPolonsky](https://github.com/BorisPolonsky/dify-helm)

 ### Configuration

@@ -120,6 +123,10 @@ For those who'd like to contribute code, see our [Contribution Guide](https://gi

 At the same time, please consider supporting Dify by sharing it on social media and at events and conferences.

+### Projects made by community
+
+- [Chatbot Chrome Extension by @charli117](https://github.com/langgenius/chatbot-chrome-extension)
+
 ### Contributors

 <a href="https://github.com/langgenius/dify/graphs/contributors">
@@ -128,25 +135,22 @@ At the same time, please consider supporting Dify by sharing it on social media

 ### Translations

-We are looking for contributors to help with translating Dify to languages other than Mandarin or English. If you are interested in helping, please see the [i18n README](https://github.com/langgenius/dify/blob/main/web/i18n/README_EN.md) for more information, and leave us a comment in the `global-users` channel of our [Discord Community Server](https://discord.gg/AhzKf7dNgk).
+We are looking for contributors to help with translating Dify to languages other than Mandarin or English. If you are interested in helping, please see the [i18n README](https://github.com/langgenius/dify/blob/main/web/i18n/README.md) for more information, and leave us a comment in the `global-users` channel of our [Discord Community Server](https://discord.gg/8Tpq4AcN9c).

 ## Community & Support

-* [Canny](https://feedback.dify.ai/). Best for: sharing feedback and checking out our feature roadmap.
+* [Github Discussion](https://github.com/langgenius/dify/discussions). Best for: sharing feedback and asking questions.
 * [GitHub Issues](https://github.com/langgenius/dify/issues). Best for: bugs you encounter using Dify.AI, and feature proposals. See our [Contribution Guide](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md).
 * [Email Support](mailto:[email protected]?subject=[GitHub]Questions%20About%20Dify). Best for: questions you have about using Dify.AI.
 * [Discord](https://discord.gg/FngNHpbcY7). Best for: sharing your applications and hanging out with the community.
 * [Twitter](https://twitter.com/dify_ai). Best for: sharing your applications and hanging out with the community.
-* [Business Contact](mailto:[email protected]?subject=[GitHub]Business%20License%20Inquiry). Best for: business inquiries of licensing Dify.AI for commercial use.

 ### Direct Meetings

-**Help us make Dify better. Reach out directly to us**.
-
 |                       Point of Contact                       |                           Purpose                            |
 | :----------------------------------------------------------: | :----------------------------------------------------------: |
-| <a href='https://cal.com/guchenhe/15min' target='_blank'><img src='https://i.postimg.cc/fWBqSmjP/Git-Hub-README-Button-3x.png' border='0' alt='Git-Hub-README-Button-3x' height="60" width="214"/></a> | Product design feedback, user experience discussions, feature planning and roadmaps. |
-| <a href='https://cal.com/pinkbanana' target='_blank'><img src='https://i.postimg.cc/LsRTh87D/Git-Hub-README-Button-2x.png' border='0' alt='Git-Hub-README-Button-2x' height="60" width="225"/></a> |        Technical support, issues, or feature requests        |
+| <a href='https://cal.com/guchenhe/15min' target='_blank'><img src='https://i.postimg.cc/fWBqSmjP/Git-Hub-README-Button-3x.png' border='0' alt='Git-Hub-README-Button-3x' height="60" width="214"/></a> | Business enquiries & product feedback. |
+| <a href='https://cal.com/pinkbanana' target='_blank'><img src='https://i.postimg.cc/LsRTh87D/Git-Hub-README-Button-2x.png' border='0' alt='Git-Hub-README-Button-2x' height="60" width="225"/></a> | Contributions, issues & feature requests |

 ## Security Disclosure

@@ -94,10 +94,12 @@ docker compose up -d

 运行后，可以在浏览器上访问 [http://localhost/install](http://localhost/install) 进入 Dify 控制台并开始初始化安装操作。

-### Helm Chart
+#### 使用 Helm Chart 部署

-非常感谢 @BorisPolonsky 为我们提供了一个 [Helm Chart](https://helm.sh/) 版本，可以在 Kubernetes 上部署 Dify。
-您可以前往 https://github.com/BorisPolonsky/dify-helm 来获取部署信息。
+使用 [Helm Chart](https://helm.sh/) 版本，可以在 Kubernetes 上部署 Dify。
+
+- [Helm Chart by @LeoQuote](https://github.com/douban/charts/tree/master/charts/dify)
+- [Helm Chart by @BorisPolonsky](https://github.com/BorisPolonsky/dify-helm)

 ### 配置

@@ -112,6 +114,7 @@ docker compose up -d

 我们欢迎您为 Dify 做出贡献，以帮助改善 Dify。包括：提交代码、问题、新想法，或分享您基于 Dify 创建的有趣且有用的 AI 应用程序。同时，我们也欢迎您在不同的活动、会议和社交媒体上分享 Dify。

+- [Github Discussion](https://github.com/langgenius/dify/discussions). 👉：分享您的应用程序并与社区交流。
 - [GitHub Issues](https://github.com/langgenius/dify/issues)。👉：使用 Dify.AI 时遇到的错误和问题，请参阅[贡献指南](CONTRIBUTING.md)。
 - [电子邮件支持](mailto:[email protected]?subject=[GitHub]Questions%20About%20Dify)。👉：关于使用 Dify.AI 的问题。
 - [Discord](https://discord.gg/FngNHpbcY7)。👉：分享您的应用程序并与社区交流。
@@ -115,6 +115,7 @@ docker compose up -d

 Difyに貢献していただき、コードの提出、問題の報告、新しいアイデアの提供、またはDifyを基に作成した興味深く有用なAIアプリケーションの共有により、Difyをより良いものにするお手伝いを歓迎します。同時に、さまざまなイベント、会議、ソーシャルメディアでDifyを共有することも歓迎します。

+- [Github Discussion](https://github.com/langgenius/dify/discussions). 👉：アプリを共有し、コミュニティとコミュニケーション。
 - [GitHub Issues](https://github.com/langgenius/dify/issues)。最適な使用法：Dify.AIの使用中に遭遇するバグやエラー、[貢献ガイド](CONTRIBUTING.md)を参照。
 - [Email サポート](mailto:[email protected]?subject=[GitHub]Questions%20About%20Dify)。最適な使用法：Dify.AIの使用に関する質問。
 - [Discord](https://discord.gg/FngNHpbcY7)。最適な使用法：アプリケーションの共有とコミュニティとの交流。
@@ -137,7 +137,15 @@ SSRF_PROXY_HTTP_URL=
 SSRF_PROXY_HTTPS_URL=

 BATCH_UPLOAD_LIMIT=10
+KEYWORD_DATA_SOURCE_TYPE=database

 # CODE EXECUTION CONFIGURATION
 CODE_EXECUTION_ENDPOINT=http://127.0.0.1:8194
 CODE_EXECUTION_API_KEY=dify-sandbox
+CODE_MAX_NUMBER=9223372036854775807
+CODE_MIN_NUMBER=-9223372036854775808
+CODE_MAX_STRING_LENGTH=80000
+TEMPLATE_TRANSFORM_MAX_LENGTH=80000
+CODE_MAX_STRING_ARRAY_LENGTH=30
+CODE_MAX_OBJECT_ARRAY_LENGTH=30
+CODE_MAX_NUMBER_ARRAY_LENGTH=1000
@@ -6,7 +6,7 @@
    "configurations": [
        {
            "name": "Python: Celery",
-            "type": "python",
+            "type": "debugpy",
            "request": "launch",
            "module": "celery",
            "justMyCode": true,
@@ -21,7 +21,7 @@
        },
        {
            "name": "Python: Flask",
-            "type": "python",
+            "type": "debugpy",
            "request": "launch",
            "module": "flask",
            "env": {
@@ -17,16 +17,16 @@
   ```bash
   sed -i "/^SECRET_KEY=/c\SECRET_KEY=$(openssl rand -base64 42)" .env
   ```
-3.5 If you use annaconda, create a new environment and activate it
+4. If you use Anaconda, create a new environment and activate it
   ```bash
   conda create --name dify python=3.10
   conda activate dify
   ```
-4. Install dependencies
+5. Install dependencies
   ```bash
   pip install -r requirements.txt
   ```
-5. Run migrate
+6. Run migrate

   Before the first launch, migrate the database to the latest version.

@@ -47,9 +47,11 @@
   pip install -r requirements.txt --upgrade --force-reinstall
   ```

-6. Start backend:
+7. Start backend:
   ```bash
   flask run --host 0.0.0.0 --port=5001 --debug
   ```
-7. Setup your application by visiting http://localhost:5001/console/api/setup or other apis...
-8. If you need to debug local async processing, you can run `celery -A app.celery worker -P gevent -c 1 --loglevel INFO -Q dataset,generation,mail`, celery can do dataset importing and other async tasks.
+8. Setup your application by visiting http://localhost:5001/console/api/setup or other apis...
+9. If you need to debug local async processing, please start the worker service by running 
+`celery -A app.celery worker -P gevent -c 1 --loglevel INFO -Q dataset,generation,mail`.
+The started celery app handles the async tasks, e.g. dataset importing and documents indexing.
@@ -109,19 +109,20 @@ def reset_encrypt_key_pair():
        click.echo(click.style('Sorry, only support SELF_HOSTED mode.', fg='red'))
        return

-    tenant = db.session.query(Tenant).first()
-    if not tenant:
-        click.echo(click.style('Sorry, no workspace found. Please enter /install to initialize.', fg='red'))
-        return
+    tenants = db.session.query(Tenant).all()
+    for tenant in tenants:
+        if not tenant:
+            click.echo(click.style('Sorry, no workspace found. Please enter /install to initialize.', fg='red'))
+            return

-    tenant.encrypt_public_key = generate_key_pair(tenant.id)
+        tenant.encrypt_public_key = generate_key_pair(tenant.id)

-    db.session.query(Provider).filter(Provider.provider_type == 'custom').delete()
-    db.session.query(ProviderModel).delete()
-    db.session.commit()
+        db.session.query(Provider).filter(Provider.provider_type == 'custom', Provider.tenant_id == tenant.id).delete()
+        db.session.query(ProviderModel).filter(ProviderModel.tenant_id == tenant.id).delete()
+        db.session.commit()

-    click.echo(click.style('Congratulations! '
-                           'the asymmetric key pair of workspace {} has been reset.'.format(tenant.id), fg='green'))
+        click.echo(click.style('Congratulations! '
+                               'the asymmetric key pair of workspace {} has been reset.'.format(tenant.id), fg='green'))


@click.command('vdb-migrate', help='migrate vector db.')
@@ -383,9 +384,17 @@ def convert_to_agent_apps():
        # fetch first 1000 apps
        sql_query = """SELECT a.id AS id FROM apps a
            INNER JOIN app_model_configs am ON a.app_model_config_id=am.id
-            WHERE a.mode = 'chat' AND am.agent_mode is not null 
-            and (am.agent_mode like '%"strategy": "function_call"%' or am.agent_mode  like '%"strategy": "react"%') 
-            and am.agent_mode like '{"enabled": true%' ORDER BY a.created_at DESC LIMIT 1000"""
+            WHERE a.mode = 'chat' 
+            AND am.agent_mode is not null 
+            AND (
+				am.agent_mode like '%"strategy": "function_call"%' 
+                OR am.agent_mode  like '%"strategy": "react"%'
+			) 
+            AND (
+				am.agent_mode like '{"enabled": true%' 
+                OR am.agent_mode like '{"max_iteration": %'
+			) ORDER BY a.created_at DESC LIMIT 1000
+        """

        with db.engine.begin() as conn:
            rs = conn.execute(db.text(sql_query))
@@ -22,6 +22,7 @@ DEFAULTS = {
    'SERVICE_API_URL': 'https://api.dify.ai',
    'APP_WEB_URL': 'https://udify.app',
    'FILES_URL': '',
+    'S3_ADDRESS_STYLE': 'auto',
    'STORAGE_TYPE': 'local',
    'STORAGE_LOCAL_PATH': 'storage',
    'CHECK_UPDATE_URL': 'https://updates.dify.ai',
@@ -64,7 +65,10 @@ DEFAULTS = {
    'KEYWORD_STORE': 'jieba',
    'BATCH_UPLOAD_LIMIT': 20,
    'CODE_EXECUTION_ENDPOINT': '',
-    'CODE_EXECUTION_API_KEY': ''
+    'CODE_EXECUTION_API_KEY': '',
+    'TOOL_ICON_CACHE_MAX_AGE': 3600,
+    'MILVUS_DATABASE': 'default',
+    'KEYWORD_DATA_SOURCE_TYPE': 'database',
 }


@@ -95,7 +99,7 @@ class Config:
        # ------------------------
        # General Configurations.
        # ------------------------
-        self.CURRENT_VERSION = "0.5.10"
+        self.CURRENT_VERSION = "0.6.1"
        self.COMMIT_SHA = get_env('COMMIT_SHA')
        self.EDITION = "SELF_HOSTED"
        self.DEPLOY_ENV = get_env('DEPLOY_ENV')
@@ -186,6 +190,7 @@ class Config:
        self.S3_ACCESS_KEY = get_env('S3_ACCESS_KEY')
        self.S3_SECRET_KEY = get_env('S3_SECRET_KEY')
        self.S3_REGION = get_env('S3_REGION')
+        self.S3_ADDRESS_STYLE = get_env('S3_ADDRESS_STYLE')
        self.AZURE_BLOB_ACCOUNT_NAME = get_env('AZURE_BLOB_ACCOUNT_NAME')
        self.AZURE_BLOB_ACCOUNT_KEY = get_env('AZURE_BLOB_ACCOUNT_KEY')
        self.AZURE_BLOB_CONTAINER_NAME = get_env('AZURE_BLOB_CONTAINER_NAME')
@@ -208,6 +213,7 @@ class Config:
        self.MILVUS_USER = get_env('MILVUS_USER')
        self.MILVUS_PASSWORD = get_env('MILVUS_PASSWORD')
        self.MILVUS_SECURE = get_env('MILVUS_SECURE')
+        self.MILVUS_DATABASE = get_env('MILVUS_DATABASE')

        # weaviate settings
        self.WEAVIATE_ENDPOINT = get_env('WEAVIATE_ENDPOINT')
@@ -311,7 +317,9 @@ class Config:
        self.CODE_EXECUTION_API_KEY = get_env('CODE_EXECUTION_API_KEY')

        self.API_COMPRESSION_ENABLED = get_bool_env('API_COMPRESSION_ENABLED')
+        self.TOOL_ICON_CACHE_MAX_AGE = get_env('TOOL_ICON_CACHE_MAX_AGE')

+        self.KEYWORD_DATA_SOURCE_TYPE = get_env('KEYWORD_DATA_SOURCE_TYPE')

 class CloudEditionConfig(Config):

@@ -8,7 +8,7 @@ api = ExternalApi(bp)
 from . import admin, apikey, extension, feature, setup, version, ping
 # Import app controllers
 from .app import (advanced_prompt_template, annotation, app, audio, completion, conversation, generator, message,
-                  model_config, site, statistic, workflow, workflow_run, workflow_app_log, workflow_statistic)
+                  model_config, site, statistic, workflow, workflow_run, workflow_app_log, workflow_statistic, agent)
 # Import auth controllers
 from .auth import activate, data_source_oauth, login, oauth
 # Import billing controllers
@@ -0,0 +1,32 @@
+from flask_restful import Resource, reqparse
+
+from controllers.console import api
+from controllers.console.app.wraps import get_app_model
+from controllers.console.setup import setup_required
+from controllers.console.wraps import account_initialization_required
+from libs.helper import uuid_value
+from libs.login import login_required
+from models.model import AppMode
+from services.agent_service import AgentService
+
+
+class AgentLogApi(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    @get_app_model(mode=[AppMode.AGENT_CHAT])
+    def get(self, app_model):
+        """Get agent logs"""
+        parser = reqparse.RequestParser()
+        parser.add_argument('message_id', type=uuid_value, required=True, location='args')
+        parser.add_argument('conversation_id', type=uuid_value, required=True, location='args')
+
+        args = parser.parse_args()
+
+        return AgentService.get_agent_logs(
+            app_model,
+            args['conversation_id'],
+            args['message_id']
+        )
+    
+api.add_resource(AgentLogApi, '/apps/<uuid:app_id>/agent/logs')
@@ -123,7 +123,6 @@ class AppApi(Resource):
                    tool_runtime = ToolManager.get_agent_tool_runtime(
                        tenant_id=current_user.current_tenant_id,
                        agent_tool=agent_tool_entity,
-                        agent_callback=None
                    )
                    manager = ToolParameterConfigurationManager(
                        tenant_id=current_user.current_tenant_id,
@@ -12,6 +12,7 @@ from controllers.console import api
 from controllers.console.app.wraps import get_app_model
 from controllers.console.setup import setup_required
 from controllers.console.wraps import account_initialization_required
+from core.app.entities.app_invoke_entities import InvokeFrom
 from extensions.ext_database import db
 from fields.conversation_fields import (
    conversation_detail_fields,
@@ -202,7 +203,7 @@ class ChatConversationApi(Resource):
            )

        if app_model.mode == AppMode.ADVANCED_CHAT.value:
-            query = query.where(Conversation.override_model_configs.is_(None))
+            query = query.where(Conversation.invoke_from != InvokeFrom.DEBUGGER.value)

        query = query.order_by(Conversation.created_at.desc())

@@ -58,7 +58,6 @@ class ModelConfigResource(Resource):
                    tool_runtime = ToolManager.get_agent_tool_runtime(
                        tenant_id=current_user.current_tenant_id,
                        agent_tool=agent_tool_entity,
-                        agent_callback=None
                    )
                    manager = ToolParameterConfigurationManager(
                        tenant_id=current_user.current_tenant_id,
@@ -96,7 +95,6 @@ class ModelConfigResource(Resource):
                        tool_runtime = ToolManager.get_agent_tool_runtime(
                            tenant_id=current_user.current_tenant_id,
                            agent_tool=agent_tool_entity,
-                            agent_callback=None
                        )
                    except Exception as e:
                        continue
@@ -1,6 +1,7 @@
 import json
 import logging

+from flask import abort, request
 from flask_restful import Resource, marshal_with, reqparse
 from werkzeug.exceptions import InternalServerError, NotFound

@@ -52,10 +53,30 @@ class DraftWorkflowApi(Resource):
        """
        Sync draft workflow
        """
-        parser = reqparse.RequestParser()
-        parser.add_argument('graph', type=dict, required=True, nullable=False, location='json')
-        parser.add_argument('features', type=dict, required=True, nullable=False, location='json')
-        args = parser.parse_args()
+        content_type = request.headers.get('Content-Type')
+
+        if 'application/json' in content_type:
+            parser = reqparse.RequestParser()
+            parser.add_argument('graph', type=dict, required=True, nullable=False, location='json')
+            parser.add_argument('features', type=dict, required=True, nullable=False, location='json')
+            args = parser.parse_args()
+        elif 'text/plain' in content_type:
+            try:
+                data = json.loads(request.data.decode('utf-8'))
+                if 'graph' not in data or 'features' not in data:
+                    raise ValueError('graph or features not found in data')
+
+                if not isinstance(data.get('graph'), dict) or not isinstance(data.get('features'), dict):
+                    raise ValueError('graph or features is not a dict')
+
+                args = {
+                    'graph': data.get('graph'),
+                    'features': data.get('features')
+                }
+            except json.JSONDecodeError:
+                return {'message': 'Invalid JSON data'}, 400
+        else:
+            abort(415)

        workflow_service = WorkflowService()
        workflow = workflow_service.sync_draft_workflow(
@@ -268,11 +289,21 @@ class ConvertToWorkflowApi(Resource):
        Convert expert mode of chatbot app to workflow mode
        Convert Completion App to Workflow App
        """
+        if request.data:
+            parser = reqparse.RequestParser()
+            parser.add_argument('name', type=str, required=False, nullable=True, location='json')
+            parser.add_argument('icon', type=str, required=False, nullable=True, location='json')
+            parser.add_argument('icon_background', type=str, required=False, nullable=True, location='json')
+            args = parser.parse_args()
+        else:
+            args = {}
+
        # convert to workflow mode
        workflow_service = WorkflowService()
        new_app_model = workflow_service.convert_to_workflow(
            app_model=app_model,
-            account=current_user
+            account=current_user,
+            args=args
        )

        # return app id
@@ -12,7 +12,11 @@ from controllers.console import api
 from controllers.console.app.error import ProviderNotInitializeError
 from controllers.console.datasets.error import InvalidActionError, NoFileUploadedError, TooManyFilesError
 from controllers.console.setup import setup_required
-from controllers.console.wraps import account_initialization_required, cloud_edition_billing_resource_check
+from controllers.console.wraps import (
+    account_initialization_required,
+    cloud_edition_billing_knowledge_limit_check,
+    cloud_edition_billing_resource_check,
+)
 from core.errors.error import LLMBadRequestError, ProviderTokenNotInitError
 from core.model_manager import ModelManager
 from core.model_runtime.entities.model_entities import ModelType
@@ -207,6 +211,7 @@ class DatasetDocumentSegmentAddApi(Resource):
    @login_required
    @account_initialization_required
    @cloud_edition_billing_resource_check('vector_space')
+    @cloud_edition_billing_knowledge_limit_check('add_segment')
    def post(self, dataset_id, document_id):
        # check dataset
        dataset_id = str(dataset_id)
@@ -357,6 +362,7 @@ class DatasetDocumentSegmentBatchImportApi(Resource):
    @login_required
    @account_initialization_required
    @cloud_edition_billing_resource_check('vector_space')
+    @cloud_edition_billing_knowledge_limit_check('add_segment')
    def post(self, dataset_id, document_id):
        # check dataset
        dataset_id = str(dataset_id)
@@ -6,6 +6,7 @@ from werkzeug.exceptions import NotFound
 from controllers.console import api
 from controllers.console.explore.error import NotChatAppError
 from controllers.console.explore.wraps import InstalledAppResource
+from core.app.entities.app_invoke_entities import InvokeFrom
 from fields.conversation_fields import conversation_infinite_scroll_pagination_fields, simple_conversation_fields
 from libs.helper import uuid_value
 from models.model import AppMode
@@ -39,8 +40,8 @@ class ConversationListApi(InstalledAppResource):
                user=current_user,
                last_id=args['last_id'],
                limit=args['limit'],
+                invoke_from=InvokeFrom.EXPLORE,
                pinned=pinned,
-                exclude_debug_conversation=True
            )
        except LastConversationNotExistsError:
            raise NotFound("Last Conversation Not Exists.")
@@ -1,6 +1,6 @@
 import io

-from flask import send_file
+from flask import current_app, send_file
 from flask_login import current_user
 from flask_restful import Resource, reqparse
 from werkzeug.exceptions import Forbidden
@@ -76,12 +76,27 @@ class ToolBuiltinProviderUpdateApi(Resource):
            provider,
            args['credentials'],
        )
+    
+class ToolBuiltinProviderGetCredentialsApi(Resource):
+    @setup_required
+    @login_required
+    @account_initialization_required
+    def get(self, provider):
+        user_id = current_user.id
+        tenant_id = current_user.current_tenant_id
+
+        return ToolManageService.get_builtin_tool_provider_credentials(
+            user_id,
+            tenant_id,
+            provider,
+        )

 class ToolBuiltinProviderIconApi(Resource):
    @setup_required
    def get(self, provider):
-        icon_bytes, minetype = ToolManageService.get_builtin_tool_provider_icon(provider)
-        return send_file(io.BytesIO(icon_bytes), mimetype=minetype)
+        icon_bytes, mimetype = ToolManageService.get_builtin_tool_provider_icon(provider)
+        icon_cache_max_age = int(current_app.config.get('TOOL_ICON_CACHE_MAX_AGE'))
+        return send_file(io.BytesIO(icon_bytes), mimetype=mimetype, max_age=icon_cache_max_age)

 class ToolModelProviderIconApi(Resource):
    @setup_required
@@ -332,6 +347,7 @@ api.add_resource(ToolProviderListApi, '/workspaces/current/tool-providers')
 api.add_resource(ToolBuiltinProviderListToolsApi, '/workspaces/current/tool-provider/builtin/<provider>/tools')
 api.add_resource(ToolBuiltinProviderDeleteApi, '/workspaces/current/tool-provider/builtin/<provider>/delete')
 api.add_resource(ToolBuiltinProviderUpdateApi, '/workspaces/current/tool-provider/builtin/<provider>/update')
+api.add_resource(ToolBuiltinProviderGetCredentialsApi, '/workspaces/current/tool-provider/builtin/<provider>/credentials')
 api.add_resource(ToolBuiltinProviderCredentialsSchemaApi, '/workspaces/current/tool-provider/builtin/<provider>/credentials_schema')
 api.add_resource(ToolBuiltinProviderIconApi, '/workspaces/current/tool-provider/builtin/<provider>/icon')
 api.add_resource(ToolModelProviderIconApi, '/workspaces/current/tool-provider/model/<provider>/icon')
@@ -339,7 +355,7 @@ api.add_resource(ToolModelProviderListToolsApi, '/workspaces/current/tool-provid
 api.add_resource(ToolApiProviderAddApi, '/workspaces/current/tool-provider/api/add')
 api.add_resource(ToolApiProviderGetRemoteSchemaApi, '/workspaces/current/tool-provider/api/remote')
 api.add_resource(ToolApiProviderListToolsApi, '/workspaces/current/tool-provider/api/tools')
-api.add_resource(ToolApiProviderUpdateApi, '/workspaces/current/tool-provider/api/update') 
+api.add_resource(ToolApiProviderUpdateApi, '/workspaces/current/tool-provider/api/update')
 api.add_resource(ToolApiProviderDeleteApi, '/workspaces/current/tool-provider/api/delete')
 api.add_resource(ToolApiProviderGetApi, '/workspaces/current/tool-provider/api/get')
 api.add_resource(ToolApiProviderSchemaApi, '/workspaces/current/tool-provider/api/schema')
@@ -51,14 +51,12 @@ def cloud_edition_billing_resource_check(resource: str,
        @wraps(view)
        def decorated(*args, **kwargs):
            features = FeatureService.get_features(current_user.current_tenant_id)
-
            if features.billing.enabled:
                members = features.members
                apps = features.apps
                vector_space = features.vector_space
                documents_upload_quota = features.documents_upload_quota
                annotation_quota_limit = features.annotation_quota_limit
-
                if resource == 'members' and 0 < members.limit <= members.size:
                    abort(403, error_msg)
                elif resource == 'apps' and 0 < apps.limit <= apps.size:
@@ -80,7 +78,29 @@ def cloud_edition_billing_resource_check(resource: str,
                    return view(*args, **kwargs)

            return view(*args, **kwargs)
+
        return decorated
+
+    return interceptor
+
+
+def cloud_edition_billing_knowledge_limit_check(resource: str,
+                                                error_msg: str = "To unlock this feature and elevate your Dify experience, please upgrade to a paid plan."):
+    def interceptor(view):
+        @wraps(view)
+        def decorated(*args, **kwargs):
+            features = FeatureService.get_features(current_user.current_tenant_id)
+            if features.billing.enabled:
+                if resource == 'add_segment':
+                    if features.billing.subscription.plan == 'sandbox':
+                        abort(403, error_msg)
+                else:
+                    return view(*args, **kwargs)
+
+            return view(*args, **kwargs)
+
+        return decorated
+
    return interceptor


@@ -99,4 +119,5 @@ def cloud_utm_record(view):
        except Exception as e:
            pass
        return view(*args, **kwargs)
+
    return decorated
@@ -6,6 +6,7 @@ import services
 from controllers.service_api import api
 from controllers.service_api.app.error import NotChatAppError
 from controllers.service_api.wraps import FetchUserArg, WhereisUserArg, validate_app_token
+from core.app.entities.app_invoke_entities import InvokeFrom
 from fields.conversation_fields import conversation_infinite_scroll_pagination_fields, simple_conversation_fields
 from libs.helper import uuid_value
 from models.model import App, AppMode, EndUser
@@ -27,7 +28,13 @@ class ConversationApi(Resource):
        args = parser.parse_args()

        try:
-            return ConversationService.pagination_by_last_id(app_model, end_user, args['last_id'], args['limit'])
+            return ConversationService.pagination_by_last_id(
+                app_model=app_model,
+                user=end_user,
+                last_id=args['last_id'],
+                limit=args['limit'],
+                invoke_from=InvokeFrom.SERVICE_API
+            )
        except services.errors.conversation.LastConversationNotExistsError:
            raise NotFound("Last Conversation Not Exists.")

@@ -63,7 +63,9 @@ class MessageListApi(Resource):
        'feedback': fields.Nested(feedback_fields, attribute='user_feedback', allow_null=True),
        'retriever_resources': fields.List(fields.Nested(retriever_resource_fields)),
        'created_at': TimestampField,
-        'agent_thoughts': fields.List(fields.Nested(agent_thought_fields))
+        'agent_thoughts': fields.List(fields.Nested(agent_thought_fields)),
+        'status': fields.String,
+        'error': fields.String,
    }

    message_infinite_scroll_pagination_fields = {
@@ -4,7 +4,11 @@ from werkzeug.exceptions import NotFound

 from controllers.service_api import api
 from controllers.service_api.app.error import ProviderNotInitializeError
-from controllers.service_api.wraps import DatasetApiResource, cloud_edition_billing_resource_check
+from controllers.service_api.wraps import (
+    DatasetApiResource,
+    cloud_edition_billing_knowledge_limit_check,
+    cloud_edition_billing_resource_check,
+)
 from core.errors.error import LLMBadRequestError, ProviderTokenNotInitError
 from core.model_manager import ModelManager
 from core.model_runtime.entities.model_entities import ModelType
@@ -18,6 +22,7 @@ class SegmentApi(DatasetApiResource):
    """Resource for segments."""

    @cloud_edition_billing_resource_check('vector_space', 'dataset')
+    @cloud_edition_billing_knowledge_limit_check('add_segment', 'dataset')
    def post(self, tenant_id, dataset_id, document_id):
        """Create single segment."""
        # check dataset
@@ -8,7 +8,7 @@ from flask import current_app, request
 from flask_login import user_logged_in
 from flask_restful import Resource
 from pydantic import BaseModel
-from werkzeug.exceptions import NotFound, Unauthorized
+from werkzeug.exceptions import Forbidden, NotFound, Unauthorized

 from extensions.ext_database import db
 from libs.login import _get_user
@@ -92,13 +92,13 @@ def cloud_edition_billing_resource_check(resource: str,
                documents_upload_quota = features.documents_upload_quota

                if resource == 'members' and 0 < members.limit <= members.size:
-                    raise Unauthorized(error_msg)
+                    raise Forbidden(error_msg)
                elif resource == 'apps' and 0 < apps.limit <= apps.size:
-                    raise Unauthorized(error_msg)
+                    raise Forbidden(error_msg)
                elif resource == 'vector_space' and 0 < vector_space.limit <= vector_space.size:
-                    raise Unauthorized(error_msg)
+                    raise Forbidden(error_msg)
                elif resource == 'documents' and 0 < documents_upload_quota.limit <= documents_upload_quota.size:
-                    raise Unauthorized(error_msg)
+                    raise Forbidden(error_msg)
                else:
                    return view(*args, **kwargs)

@@ -107,6 +107,27 @@ def cloud_edition_billing_resource_check(resource: str,
    return interceptor


+def cloud_edition_billing_knowledge_limit_check(resource: str,
+                                                api_token_type: str,
+                                                error_msg: str = "To unlock this feature and elevate your Dify experience, please upgrade to a paid plan."):
+    def interceptor(view):
+        @wraps(view)
+        def decorated(*args, **kwargs):
+            api_token = validate_and_get_api_token(api_token_type)
+            features = FeatureService.get_features(api_token.tenant_id)
+            if features.billing.enabled:
+                if resource == 'add_segment':
+                    if features.billing.subscription.plan == 'sandbox':
+                        raise Forbidden(error_msg)
+                else:
+                    return view(*args, **kwargs)
+
+            return view(*args, **kwargs)
+
+        return decorated
+
+    return interceptor
+
 def validate_dataset_token(view=None):
    def decorator(view):
        @wraps(view)
@@ -5,6 +5,7 @@ from werkzeug.exceptions import NotFound
 from controllers.web import api
 from controllers.web.error import NotChatAppError
 from controllers.web.wraps import WebApiResource
+from core.app.entities.app_invoke_entities import InvokeFrom
 from fields.conversation_fields import conversation_infinite_scroll_pagination_fields, simple_conversation_fields
 from libs.helper import uuid_value
 from models.model import AppMode
@@ -37,7 +38,8 @@ class ConversationListApi(WebApiResource):
                user=end_user,
                last_id=args['last_id'],
                limit=args['limit'],
-                pinned=pinned
+                invoke_from=InvokeFrom.WEB_APP,
+                pinned=pinned,
            )
        except LastConversationNotExistsError:
            raise NotFound("Last Conversation Not Exists.")
@@ -66,7 +66,9 @@ class MessageListApi(WebApiResource):
        'feedback': fields.Nested(feedback_fields, attribute='user_feedback', allow_null=True),
        'retriever_resources': fields.List(fields.Nested(retriever_resource_fields)),
        'created_at': TimestampField,
-        'agent_thoughts': fields.List(fields.Nested(agent_thought_fields))
+        'agent_thoughts': fields.List(fields.Nested(agent_thought_fields)),
+        'status': fields.String,
+        'error': fields.String,
    }

    message_infinite_scroll_pagination_fields = {
@@ -10,12 +10,10 @@ from core.app.apps.base_app_queue_manager import AppQueueManager
 from core.app.apps.base_app_runner import AppRunner
 from core.app.entities.app_invoke_entities import (
    AgentChatAppGenerateEntity,
-    InvokeFrom,
    ModelConfigWithCredentialsEntity,
 )
 from core.callback_handler.agent_tool_callback_handler import DifyAgentCallbackHandler
 from core.callback_handler.index_tool_callback_handler import DatasetIndexToolCallbackHandler
-from core.file.message_file_parser import FileTransferMethod
 from core.memory.token_buffer_memory import TokenBufferMemory
 from core.model_manager import ModelInstance
 from core.model_runtime.entities.llm_entities import LLMUsage
@@ -32,7 +30,6 @@ from core.model_runtime.model_providers.__base.large_language_model import Large
 from core.model_runtime.utils.encoders import jsonable_encoder
 from core.tools.entities.tool_entities import (
    ToolInvokeMessage,
-    ToolInvokeMessageBinary,
    ToolParameter,
    ToolRuntimeVariablePool,
 )
@@ -40,7 +37,7 @@ from core.tools.tool.dataset_retriever_tool import DatasetRetrieverTool
 from core.tools.tool.tool import Tool
 from core.tools.tool_manager import ToolManager
 from extensions.ext_database import db
-from models.model import Message, MessageAgentThought, MessageFile
+from models.model import Message, MessageAgentThought
 from models.tools import ToolConversationVariables

 logger = logging.getLogger(__name__)
@@ -143,7 +140,7 @@ class BaseAgentRunner(AppRunner):
                result += f"result link: {response.message}. please tell user to check it."
            elif response.type == ToolInvokeMessage.MessageType.IMAGE_LINK or \
                 response.type == ToolInvokeMessage.MessageType.IMAGE:
-                result += "image has been created and sent to user already, you should tell user to check it now."
+                result += "image has been created and sent to user already, you do not need to create it, just tell the user to check it now."
            else:
                result += f"tool response: {response.message}."

@@ -156,7 +153,6 @@ class BaseAgentRunner(AppRunner):
        tool_entity = ToolManager.get_agent_tool_runtime(
            tenant_id=self.tenant_id,
            agent_tool=tool,
-            agent_callback=self.agent_callback
        )
        tool_entity.load_variables(self.variables_pool)

@@ -270,87 +266,6 @@ class BaseAgentRunner(AppRunner):
                    prompt_tool.parameters['required'].append(parameter.name)

        return prompt_tool
-    
-    def extract_tool_response_binary(self, tool_response: list[ToolInvokeMessage]) -> list[ToolInvokeMessageBinary]:
-        """
-        Extract tool response binary
-        """
-        result = []
-
-        for response in tool_response:
-            if response.type == ToolInvokeMessage.MessageType.IMAGE_LINK or \
-                response.type == ToolInvokeMessage.MessageType.IMAGE:
-                result.append(ToolInvokeMessageBinary(
-                    mimetype=response.meta.get('mime_type', 'octet/stream'),
-                    url=response.message,
-                    save_as=response.save_as,
-                ))
-            elif response.type == ToolInvokeMessage.MessageType.BLOB:
-                result.append(ToolInvokeMessageBinary(
-                    mimetype=response.meta.get('mime_type', 'octet/stream'),
-                    url=response.message,
-                    save_as=response.save_as,
-                ))
-            elif response.type == ToolInvokeMessage.MessageType.LINK:
-                # check if there is a mime type in meta
-                if response.meta and 'mime_type' in response.meta:
-                    result.append(ToolInvokeMessageBinary(
-                        mimetype=response.meta.get('mime_type', 'octet/stream') if response.meta else 'octet/stream',
-                        url=response.message,
-                        save_as=response.save_as,
-                    ))
-
-        return result
-    
-    def create_message_files(self, messages: list[ToolInvokeMessageBinary]) -> list[tuple[MessageFile, bool]]:
-        """
-        Create message file
-
-        :param messages: messages
-        :return: message files, should save as variable
-        """
-        result = []
-
-        for message in messages:
-            file_type = 'bin'
-            if 'image' in message.mimetype:
-                file_type = 'image'
-            elif 'video' in message.mimetype:
-                file_type = 'video'
-            elif 'audio' in message.mimetype:
-                file_type = 'audio'
-            elif 'text' in message.mimetype:
-                file_type = 'text'
-            elif 'pdf' in message.mimetype:
-                file_type = 'pdf'
-            elif 'zip' in message.mimetype:
-                file_type = 'archive'
-            # ...
-
-            invoke_from = self.application_generate_entity.invoke_from
-
-            message_file = MessageFile(
-                message_id=self.message.id,
-                type=file_type,
-                transfer_method=FileTransferMethod.TOOL_FILE.value,
-                belongs_to='assistant',
-                url=message.url,
-                upload_file_id=None,
-                created_by_role=('account'if invoke_from in [InvokeFrom.EXPLORE, InvokeFrom.DEBUGGER] else 'end_user'),
-                created_by=self.user_id,
-            )
-            db.session.add(message_file)
-            db.session.commit()
-            db.session.refresh(message_file)
-
-            result.append((
-                message_file,
-                message.save_as
-            ))
-
-        db.session.close()
-
-        return result
        
    def create_agent_thought(self, message_id: str, message: str, 
                             tool_name: str, tool_input: str, messages_ids: list[str]
@@ -364,6 +279,7 @@ class BaseAgentRunner(AppRunner):
            thought='',
            tool=tool_name,
            tool_labels_str='{}',
+            tool_meta_str='{}',
            tool_input=tool_input,
            message=message,
            message_token=0,
@@ -398,7 +314,8 @@ class BaseAgentRunner(AppRunner):
                           tool_name: str,
                           tool_input: Union[str, dict],
                           thought: str, 
-                           observation: str, 
+                           observation: Union[str, str], 
+                           tool_invoke_meta: Union[str, dict],
                           answer: str,
                           messages_ids: list[str],
                           llm_usage: LLMUsage = None) -> MessageAgentThought:
@@ -425,6 +342,12 @@ class BaseAgentRunner(AppRunner):
            agent_thought.tool_input = tool_input

        if observation is not None:
+            if isinstance(observation, dict):
+                try:
+                    observation = json.dumps(observation, ensure_ascii=False)
+                except Exception as e:
+                    observation = json.dumps(observation)
+                    
            agent_thought.observation = observation

        if answer is not None:
@@ -458,6 +381,15 @@ class BaseAgentRunner(AppRunner):

        agent_thought.tool_labels_str = json.dumps(labels)

+        if tool_invoke_meta is not None:
+            if isinstance(tool_invoke_meta, dict):
+                try:
+                    tool_invoke_meta = json.dumps(tool_invoke_meta, ensure_ascii=False)
+                except Exception as e:
+                    tool_invoke_meta = json.dumps(tool_invoke_meta)
+
+            agent_thought.tool_meta_str = tool_invoke_meta
+
        db.session.commit()
        db.session.close()
    
@@ -497,7 +429,15 @@ class BaseAgentRunner(AppRunner):
                        tools = tools.split(';')
                        tool_calls: list[AssistantPromptMessage.ToolCall] = []
                        tool_call_response: list[ToolPromptMessage] = []
-                        tool_inputs = json.loads(agent_thought.tool_input)
+                        try:
+                            tool_inputs = json.loads(agent_thought.tool_input)
+                        except Exception as e:
+                            tool_inputs = { tool: {} for tool in tools }
+                        try:
+                            tool_responses = json.loads(agent_thought.observation)
+                        except Exception as e:
+                            tool_responses = { tool: agent_thought.observation for tool in tools }
+
                        for tool in tools:
                            # generate a uuid for tool call
                            tool_call_id = str(uuid.uuid4())
@@ -510,7 +450,7 @@ class BaseAgentRunner(AppRunner):
                                )
                            ))
                            tool_call_response.append(ToolPromptMessage(
-                                content=agent_thought.observation,
+                                content=tool_responses.get(tool, agent_thought.observation),
                                name=tool,
                                tool_call_id=tool_call_id,
                            ))
@@ -530,4 +470,4 @@ class BaseAgentRunner(AppRunner):

        db.session.close()

-        return result
+        return result
@@ -17,15 +17,8 @@ from core.model_runtime.entities.message_entities import (
    UserPromptMessage,
 )
 from core.model_runtime.utils.encoders import jsonable_encoder
-from core.tools.errors import (
-    ToolInvokeError,
-    ToolNotFoundError,
-    ToolNotSupportedError,
-    ToolParameterValidationError,
-    ToolProviderCredentialValidationError,
-    ToolProviderNotFoundError,
-)
-from core.tools.utils.message_transformer import ToolFileMessageTransformer
+from core.tools.entities.tool_entities import ToolInvokeMeta
+from core.tools.tool_engine import ToolEngine
 from models.model import Conversation, Message


@@ -191,7 +184,7 @@ class CotAgentRunner(BaseAgentRunner):
                            delta=LLMResultChunkDelta(
                                index=0,
                                message=AssistantPromptMessage(
-                                    content=json.dumps(chunk)
+                                    content=json.dumps(chunk, ensure_ascii=False) # if ensure_ascii=True, the text in webui maybe garbled text
                                ),
                                usage=None
                            )
@@ -223,7 +216,10 @@ class CotAgentRunner(BaseAgentRunner):
            
            self.save_agent_thought(agent_thought=agent_thought,
                                    tool_name=scratchpad.action.action_name if scratchpad.action else '',
-                                    tool_input=scratchpad.action.action_input if scratchpad.action else '',
+                                    tool_input={
+                                        scratchpad.action.action_name: scratchpad.action.action_input
+                                    } if scratchpad.action else '',
+                                    tool_invoke_meta={},
                                    thought=scratchpad.thought,
                                    observation='',
                                    answer=scratchpad.agent_response,
@@ -256,71 +252,65 @@ class CotAgentRunner(BaseAgentRunner):
                    tool_instance = tool_instances.get(tool_call_name)
                    if not tool_instance:
                        answer = f"there is not a tool named {tool_call_name}"
-                        self.save_agent_thought(agent_thought=agent_thought, 
-                                                tool_name='',
-                                                tool_input='',
-                                                thought=None, 
-                                                observation=answer, 
-                                                answer=answer,
-                                                messages_ids=[])
+                        self.save_agent_thought(
+                            agent_thought=agent_thought, 
+                            tool_name='',
+                            tool_input='',
+                            tool_invoke_meta=ToolInvokeMeta.error_instance(
+                                f"there is not a tool named {tool_call_name}"
+                            ).to_dict(),
+                            thought=None, 
+                            observation={
+                                tool_call_name: answer
+                            }, 
+                            answer=answer,
+                            messages_ids=[]
+                        )
                        self.queue_manager.publish(QueueAgentThoughtEvent(
                            agent_thought_id=agent_thought.id
                        ), PublishFrom.APPLICATION_MANAGER)
                    else:
+                        if isinstance(tool_call_args, str):
+                            try:
+                                tool_call_args = json.loads(tool_call_args)
+                            except json.JSONDecodeError:
+                                pass
+
                        # invoke tool
-                        error_response = None
-                        try:
-                            if isinstance(tool_call_args, str):
-                                try:
-                                    tool_call_args = json.loads(tool_call_args)
-                                except json.JSONDecodeError:
-                                    pass
+                        tool_invoke_response, message_files, tool_invoke_meta = ToolEngine.agent_invoke(
+                            tool=tool_instance,
+                            tool_parameters=tool_call_args,
+                            user_id=self.user_id,
+                            tenant_id=self.tenant_id,
+                            message=self.message,
+                            invoke_from=self.application_generate_entity.invoke_from,
+                            agent_tool_callback=self.agent_callback
+                        )
+                        # publish files
+                        for message_file, save_as in message_files:
+                            if save_as:
+                                self.variables_pool.set_file(tool_name=tool_call_name, value=message_file.id, name=save_as)

-                            tool_response = tool_instance.invoke(
-                                user_id=self.user_id, 
-                                tool_parameters=tool_call_args
-                            )
-                            # transform tool response to llm friendly response
-                            tool_response = ToolFileMessageTransformer.transform_tool_invoke_messages(
-                                messages=tool_response, 
-                                user_id=self.user_id, 
-                                tenant_id=self.tenant_id, 
-                                conversation_id=self.message.conversation_id
-                            )
-                            # extract binary data from tool invoke message
-                            binary_files = self.extract_tool_response_binary(tool_response)
-                            # create message file
-                            message_files = self.create_message_files(binary_files)
-                            # publish files
-                            for message_file, save_as in message_files:
-                                if save_as:
-                                    self.variables_pool.set_file(tool_name=tool_call_name,
-                                                                  value=message_file.id,
-                                                                  name=save_as)
-                                self.queue_manager.publish(QueueMessageFileEvent(
-                                    message_file_id=message_file.id
-                                ), PublishFrom.APPLICATION_MANAGER)
+                            # publish message file
+                            self.queue_manager.publish(QueueMessageFileEvent(
+                                message_file_id=message_file.id
+                            ), PublishFrom.APPLICATION_MANAGER)
+                            # add message file ids
+                            message_file_ids.append(message_file.id)

-                            message_file_ids = [message_file.id for message_file, _ in message_files]
-                        except ToolProviderCredentialValidationError as e:
-                            error_response = "Please check your tool provider credentials"
-                        except (
-                            ToolNotFoundError, ToolNotSupportedError, ToolProviderNotFoundError
-                        ) as e:
-                            error_response = f"there is not a tool named {tool_call_name}"
-                        except (
-                            ToolParameterValidationError
-                        ) as e:
-                            error_response = f"tool parameters validation error: {e}, please check your tool parameters"
-                        except ToolInvokeError as e:
-                            error_response = f"tool invoke error: {e}"
-                        except Exception as e:
-                            error_response = f"unknown error: {e}"
+                        # publish files
+                        for message_file, save_as in message_files:
+                            if save_as:
+                                self.variables_pool.set_file(tool_name=tool_call_name,
+                                                                value=message_file.id,
+                                                                name=save_as)
+                            self.queue_manager.publish(QueueMessageFileEvent(
+                                message_file_id=message_file.id
+                            ), PublishFrom.APPLICATION_MANAGER)

-                        if error_response:
-                            observation = error_response
-                        else:
-                            observation = self._convert_tool_response_to_str(tool_response)
+                        message_file_ids = [message_file.id for message_file, _ in message_files]
+
+                        observation = tool_invoke_response

                        # save scratchpad
                        scratchpad.observation = observation
@@ -329,9 +319,16 @@ class CotAgentRunner(BaseAgentRunner):
                        self.save_agent_thought(
                            agent_thought=agent_thought, 
                            tool_name=tool_call_name,
-                            tool_input=tool_call_args,
+                            tool_input={
+                                tool_call_name: tool_call_args
+                            },
+                            tool_invoke_meta={
+                                tool_call_name: tool_invoke_meta.to_dict()
+                            },
                            thought=None,
-                            observation=observation, 
+                            observation={
+                                tool_call_name: observation
+                            }, 
                            answer=scratchpad.agent_response,
                            messages_ids=message_file_ids,
                        )
@@ -362,9 +359,10 @@ class CotAgentRunner(BaseAgentRunner):
        self.save_agent_thought(
            agent_thought=agent_thought, 
            tool_name='',
-            tool_input='',
+            tool_input={},
+            tool_invoke_meta={},
            thought=final_answer,
-            observation='', 
+            observation={}, 
            answer=final_answer,
            messages_ids=[]
        )
@@ -15,15 +15,8 @@ from core.model_runtime.entities.message_entities import (
    ToolPromptMessage,
    UserPromptMessage,
 )
-from core.tools.errors import (
-    ToolInvokeError,
-    ToolNotFoundError,
-    ToolNotSupportedError,
-    ToolParameterValidationError,
-    ToolProviderCredentialValidationError,
-    ToolProviderNotFoundError,
-)
-from core.tools.utils.message_transformer import ToolFileMessageTransformer
+from core.tools.entities.tool_entities import ToolInvokeMeta
+from core.tools.tool_engine import ToolEngine
 from models.model import Conversation, Message, MessageAgentThought

 logger = logging.getLogger(__name__)
@@ -214,19 +207,25 @@ class FunctionCallAgentRunner(BaseAgentRunner):
                    )
                )

+            assistant_message = AssistantPromptMessage(
+                content='',
+                tool_calls=[]
+            )
            if tool_calls:
-                prompt_messages.append(AssistantPromptMessage(
-                    content='',
-                    name='',
-                    tool_calls=[AssistantPromptMessage.ToolCall(
+                assistant_message.tool_calls=[
+                    AssistantPromptMessage.ToolCall(
                        id=tool_call[0],
                        type='function',
                        function=AssistantPromptMessage.ToolCall.ToolCallFunction(
                            name=tool_call[1],
                            arguments=json.dumps(tool_call[2], ensure_ascii=False)
                        )
-                    ) for tool_call in tool_calls]
-                ))
+                    ) for tool_call in tool_calls
+                ]
+            else:
+                assistant_message.content = response
+            
+            prompt_messages.append(assistant_message)

            # save thought
            self.save_agent_thought(
@@ -234,6 +233,7 @@ class FunctionCallAgentRunner(BaseAgentRunner):
                tool_name=tool_call_names,
                tool_input=tool_call_inputs,
                thought=response,
+                tool_invoke_meta=None,
                observation=None,
                answer=response,
                messages_ids=[],
@@ -245,12 +245,6 @@ class FunctionCallAgentRunner(BaseAgentRunner):
            
            final_answer += response + '\n'

-            # update prompt messages
-            if response.strip():
-                prompt_messages.append(AssistantPromptMessage(
-                    content=response,
-                ))
-            
            # call tools
            tool_responses = []
            for tool_call_id, tool_call_name, tool_call_args in tool_calls:
@@ -259,72 +253,40 @@ class FunctionCallAgentRunner(BaseAgentRunner):
                    tool_response = {
                        "tool_call_id": tool_call_id,
                        "tool_call_name": tool_call_name,
-                        "tool_response": f"there is not a tool named {tool_call_name}"
+                        "tool_response": f"there is not a tool named {tool_call_name}",
+                        "meta": ToolInvokeMeta.error_instance(f"there is not a tool named {tool_call_name}").to_dict()
                    }
-                    tool_responses.append(tool_response)
                else:
                    # invoke tool
-                    error_response = None
-                    try:
-                        tool_invoke_message = tool_instance.invoke(
-                            user_id=self.user_id, 
-                            tool_parameters=tool_call_args, 
-                        )
-                        # transform tool invoke message to get LLM friendly message
-                        tool_invoke_message = ToolFileMessageTransformer.transform_tool_invoke_messages(
-                            messages=tool_invoke_message, 
-                            user_id=self.user_id, 
-                            tenant_id=self.tenant_id, 
-                            conversation_id=self.message.conversation_id
-                        )
-                        # extract binary data from tool invoke message
-                        binary_files = self.extract_tool_response_binary(tool_invoke_message)
-                        # create message file
-                        message_files = self.create_message_files(binary_files)
-                        # publish files
-                        for message_file, save_as in message_files:
-                            if save_as:
-                                self.variables_pool.set_file(tool_name=tool_call_name, value=message_file.id, name=save_as)
-
-                            # publish message file
-                            self.queue_manager.publish(QueueMessageFileEvent(
-                                message_file_id=message_file.id
-                            ), PublishFrom.APPLICATION_MANAGER)
-                            # add message file ids
-                            message_file_ids.append(message_file.id)
-                            
-                    except ToolProviderCredentialValidationError as e:
-                        error_response = "Please check your tool provider credentials"
-                    except (
-                        ToolNotFoundError, ToolNotSupportedError, ToolProviderNotFoundError
-                    ) as e:
-                        error_response = f"there is not a tool named {tool_call_name}"
-                    except (
-                        ToolParameterValidationError
-                    ) as e:
-                        error_response = f"tool parameters validation error: {e}, please check your tool parameters"
-                    except ToolInvokeError as e:
-                        error_response = f"tool invoke error: {e}"
-                    except Exception as e:
-                        error_response = f"unknown error: {e}"
-
-                    if error_response:
-                        observation = error_response
-                        tool_response = {
-                            "tool_call_id": tool_call_id,
-                            "tool_call_name": tool_call_name,
-                            "tool_response": error_response
-                        }
-                        tool_responses.append(tool_response)
-                    else:
-                        observation = self._convert_tool_response_to_str(tool_invoke_message)
-                        tool_response = {
-                            "tool_call_id": tool_call_id,
-                            "tool_call_name": tool_call_name,
-                            "tool_response": observation
-                        }
-                        tool_responses.append(tool_response)
+                    tool_invoke_response, message_files, tool_invoke_meta = ToolEngine.agent_invoke(
+                        tool=tool_instance,
+                        tool_parameters=tool_call_args,
+                        user_id=self.user_id,
+                        tenant_id=self.tenant_id,
+                        message=self.message,
+                        invoke_from=self.application_generate_entity.invoke_from,
+                        agent_tool_callback=self.agent_callback,
+                    )
+                    # publish files
+                    for message_file, save_as in message_files:
+                        if save_as:
+                            self.variables_pool.set_file(tool_name=tool_call_name, value=message_file.id, name=save_as)

+                        # publish message file
+                        self.queue_manager.publish(QueueMessageFileEvent(
+                            message_file_id=message_file.id
+                        ), PublishFrom.APPLICATION_MANAGER)
+                        # add message file ids
+                        message_file_ids.append(message_file.id)
+                    
+                    tool_response = {
+                        "tool_call_id": tool_call_id,
+                        "tool_call_name": tool_call_name,
+                        "tool_response": tool_invoke_response,
+                        "meta": tool_invoke_meta.to_dict()
+                    }
+                
+                tool_responses.append(tool_response)
                prompt_messages = self.organize_prompt_messages(
                    prompt_template=prompt_template,
                    query=None,
@@ -341,7 +303,14 @@ class FunctionCallAgentRunner(BaseAgentRunner):
                    tool_name=None,
                    tool_input=None,
                    thought=None, 
-                    observation=tool_response['tool_response'], 
+                    tool_invoke_meta={
+                        tool_response['tool_call_name']: tool_response['meta'] 
+                        for tool_response in tool_responses
+                    },
+                    observation={
+                        tool_response['tool_call_name']: tool_response['tool_response'] 
+                        for tool_response in tool_responses
+                    },
                    answer=None,
                    messages_ids=message_file_ids
                )
@@ -39,12 +39,12 @@ class SensitiveWordAvoidanceConfigManager:

            if not only_structure_validate:
                typ = config["sensitive_word_avoidance"]["type"]
-                config = config["sensitive_word_avoidance"]["config"]
+                sensitive_word_avoidance_config = config["sensitive_word_avoidance"]["config"]

                ModerationFactory.validate_config(
                    name=typ,
                    tenant_id=tenant_id,
-                    config=config
+                    config=sensitive_word_avoidance_config
                )

        return config, ["sensitive_word_avoidance"]
@@ -189,7 +189,6 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
                logger.exception("Validation Error when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except (ValueError, InvokeError) as e:
-                logger.exception("Bad Request when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except Exception as e:
                logger.exception("Unknown Error when generating")
@@ -196,6 +196,7 @@ class AdvancedChatAppGenerateTaskPipeline(BasedGenerateTaskPipeline, WorkflowCyc
                    yield from self._generate_stream_outputs_when_node_started()

                yield self._workflow_node_start_to_stream_response(
+                    event=event,
                    task_id=self._application_generate_entity.task_id,
                    workflow_node_execution=workflow_node_execution
                )
@@ -214,16 +215,16 @@ class AdvancedChatAppGenerateTaskPipeline(BasedGenerateTaskPipeline, WorkflowCyc
            elif isinstance(event, QueueStopEvent | QueueWorkflowSucceededEvent | QueueWorkflowFailedEvent):
                workflow_run = self._handle_workflow_finished(event)
                if workflow_run:
-                    if workflow_run.status == WorkflowRunStatus.FAILED.value:
-                        err_event = QueueErrorEvent(error=ValueError(f'Run failed: {workflow_run.error}'))
-                        yield self._error_to_stream_response(self._handle_error(err_event))
-                        break
-
                    yield self._workflow_finish_to_stream_response(
                        task_id=self._application_generate_entity.task_id,
                        workflow_run=workflow_run
                    )

+                    if workflow_run.status == WorkflowRunStatus.FAILED.value:
+                        err_event = QueueErrorEvent(error=ValueError(f'Run failed: {workflow_run.error}'))
+                        yield self._error_to_stream_response(self._handle_error(err_event, self._message))
+                        break
+
                if isinstance(event, QueueStopEvent):
                    # Save message
                    self._save_message()
@@ -143,6 +143,12 @@ class AgentChatAppConfigManager(BaseAppConfigManager):
        config, current_related_config_keys = RetrievalResourceConfigManager.validate_and_set_defaults(config)
        related_config_keys.extend(current_related_config_keys)

+        # dataset configs
+        # dataset_query_variable
+        config, current_related_config_keys = DatasetConfigManager.validate_and_set_defaults(tenant_id, app_mode,
+                                                                                             config)
+        related_config_keys.extend(current_related_config_keys)
+
        # moderation validation
        config, current_related_config_keys = SensitiveWordAvoidanceConfigManager.validate_and_set_defaults(tenant_id,
                                                                                                            config)
@@ -198,7 +198,6 @@ class AgentChatAppGenerator(MessageBasedAppGenerator):
                logger.exception("Validation Error when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except (ValueError, InvokeError) as e:
-                logger.exception("Bad Request when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except Exception as e:
                logger.exception("Unknown Error when generating")
@@ -169,10 +169,11 @@ class AppRunner:
                    text=advanced_completion_prompt_template.prompt
                )

-                memory_config.role_prefix = MemoryConfig.RolePrefix(
-                    user=advanced_completion_prompt_template.role_prefix.user,
-                    assistant=advanced_completion_prompt_template.role_prefix.assistant
-                )
+                if advanced_completion_prompt_template.role_prefix:
+                    memory_config.role_prefix = MemoryConfig.RolePrefix(
+                        user=advanced_completion_prompt_template.role_prefix.user,
+                        assistant=advanced_completion_prompt_template.role_prefix.assistant
+                    )
            else:
                prompt_template = []
                for message in prompt_template_entity.advanced_chat_prompt_template.messages:
@@ -195,7 +195,6 @@ class ChatAppGenerator(MessageBasedAppGenerator):
                logger.exception("Validation Error when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except (ValueError, InvokeError) as e:
-                logger.exception("Bad Request when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except Exception as e:
                logger.exception("Unknown Error when generating")
@@ -184,7 +184,6 @@ class CompletionAppGenerator(MessageBasedAppGenerator):
                logger.exception("Validation Error when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except (ValueError, InvokeError) as e:
-                logger.exception("Bad Request when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except Exception as e:
                logger.exception("Unknown Error when generating")
@@ -137,7 +137,6 @@ class WorkflowAppGenerator(BaseAppGenerator):
                logger.exception("Validation Error when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except (ValueError, InvokeError) as e:
-                logger.exception("Bad Request when generating")
                queue_manager.publish_error(e, PublishFrom.APPLICATION_MANAGER)
            except Exception as e:
                logger.exception("Unknown Error when generating")
@@ -162,6 +162,7 @@ class WorkflowAppGenerateTaskPipeline(BasedGenerateTaskPipeline, WorkflowCycleMa
            elif isinstance(event, QueueNodeStartedEvent):
                workflow_node_execution = self._handle_node_start(event)
                yield self._workflow_node_start_to_stream_response(
+                    event=event,
                    task_id=self._application_generate_entity.task_id,
                    workflow_node_execution=workflow_node_execution
                )
@@ -122,7 +122,7 @@ class MessageEndStreamResponse(StreamResponse):
    """
    event: StreamEvent = StreamEvent.MESSAGE_END
    id: str
-    metadata: Optional[dict] = None
+    metadata: dict = {}


 class MessageFileStreamResponse(StreamResponse):
@@ -179,6 +179,7 @@ class WorkflowStartStreamResponse(StreamResponse):
        id: str
        workflow_id: str
        sequence_number: int
+        inputs: dict
        created_at: int

    event: StreamEvent = StreamEvent.WORKFLOW_STARTED
@@ -229,6 +230,7 @@ class NodeStartStreamResponse(StreamResponse):
        predecessor_node_id: Optional[str] = None
        inputs: Optional[dict] = None
        created_at: int
+        extras: dict = {}

    event: StreamEvent = StreamEvent.NODE_STARTED
    workflow_run_id: str
@@ -14,7 +14,7 @@ from core.app.entities.task_entities import (
    PingStreamResponse,
    TaskState,
 )
-from core.errors.error import ModelCurrentlyNotSupportError, ProviderTokenNotInitError, QuotaExceededError
+from core.errors.error import QuotaExceededError
 from core.model_runtime.errors.invoke import InvokeAuthorizationError, InvokeError
 from core.moderation.output_moderation import ModerationRule, OutputModeration
 from extensions.ext_database import db
@@ -83,24 +83,12 @@ class BasedGenerateTaskPipeline:
        :param e: exception
        :return:
        """
-        error_responses = {
-            ValueError: None,
-            ProviderTokenNotInitError: None,
-            QuotaExceededError: "Your quota for Dify Hosted Model Provider has been exhausted. "
-                                "Please go to Settings -> Model Provider to complete your own provider credentials.",
-            ModelCurrentlyNotSupportError: None,
-            InvokeError: None
-        }
+        if isinstance(e, QuotaExceededError):
+            return ("Your quota for Dify Hosted Model Provider has been exhausted. "
+                    "Please go to Settings -> Model Provider to complete your own provider credentials.")

-        # Determine the response based on the type of exception
-        data = None
-        for k, v in error_responses.items():
-            if isinstance(e, k):
-                data = v
-
-        if data:
-            message = getattr(e, 'description', str(e)) if data is None else data
-        else:
+        message = getattr(e, 'description', str(e))
+        if not message:
            message = 'Internal Server Error, please contact support.'

        return message
@@ -1,7 +1,7 @@
 import json
 import time
 from datetime import datetime
-from typing import Any, Optional, Union
+from typing import Any, Optional, Union, cast

 from core.app.entities.app_invoke_entities import AdvancedChatAppGenerateEntity, InvokeFrom, WorkflowAppGenerateEntity
 from core.app.entities.queue_entities import (
@@ -23,7 +23,9 @@ from core.app.entities.task_entities import (
 )
 from core.file.file_obj import FileVar
 from core.model_runtime.utils.encoders import jsonable_encoder
+from core.tools.tool_manager import ToolManager
 from core.workflow.entities.node_entities import NodeRunMetadataKey, NodeType, SystemVariable
+from core.workflow.nodes.tool.entities import ToolNodeData
 from core.workflow.workflow_engine_manager import WorkflowEngineManager
 from extensions.ext_database import db
 from models.account import Account
@@ -69,6 +71,9 @@ class WorkflowCycleManage:

        inputs = {**user_inputs}
        for key, value in (system_inputs or {}).items():
+            if key.value == 'conversation':
+                continue
+
            inputs[f'sys.{key.value}'] = value
        inputs = WorkflowEngineManager.handle_special_values(inputs)

@@ -272,6 +277,7 @@ class WorkflowCycleManage:
                id=workflow_run.id,
                workflow_id=workflow_run.workflow_id,
                sequence_number=workflow_run.sequence_number,
+                inputs=workflow_run.inputs_dict,
                created_at=int(workflow_run.created_at.timestamp())
            )
        )
@@ -321,15 +327,18 @@ class WorkflowCycleManage:
            )
        )

-    def _workflow_node_start_to_stream_response(self, task_id: str, workflow_node_execution: WorkflowNodeExecution) \
+    def _workflow_node_start_to_stream_response(self, event: QueueNodeStartedEvent,
+                                                task_id: str,
+                                                workflow_node_execution: WorkflowNodeExecution) \
            -> NodeStartStreamResponse:
        """
        Workflow node start to stream response.
+        :param event: queue node started event
        :param task_id: task id
        :param workflow_node_execution: workflow node execution
        :return:
        """
-        return NodeStartStreamResponse(
+        response = NodeStartStreamResponse(
            task_id=task_id,
            workflow_run_id=workflow_node_execution.workflow_run_id,
            data=NodeStartStreamResponse.Data(
@@ -344,6 +353,17 @@ class WorkflowCycleManage:
            )
        )

+        # extras logic
+        if event.node_type == NodeType.TOOL:
+            node_data = cast(ToolNodeData, event.node_data)
+            response.data.extras['icon'] = ToolManager.get_tool_icon(
+                tenant_id=self._application_generate_entity.app_config.tenant_id,
+                provider_type=node_data.provider_type,
+                provider_id=node_data.provider_id
+            )
+
+        return response
+
    def _workflow_node_finish_to_stream_response(self, task_id: str, workflow_node_execution: WorkflowNodeExecution) \
            -> NodeFinishStreamResponse:
        """
@@ -1,12 +1,32 @@
 import os
-from typing import Any, Optional, Union
+from typing import Any, Optional, TextIO, Union

-from langchain.callbacks.base import BaseCallbackHandler
-from langchain.input import print_text
 from pydantic import BaseModel

+_TEXT_COLOR_MAPPING = {
+    "blue": "36;1",
+    "yellow": "33;1",
+    "pink": "38;5;200",
+    "green": "32;1",
+    "red": "31;1",
+}

-class DifyAgentCallbackHandler(BaseCallbackHandler, BaseModel):
+def get_colored_text(text: str, color: str) -> str:
+    """Get colored text."""
+    color_str = _TEXT_COLOR_MAPPING[color]
+    return f"\u001b[{color_str}m\033[1;3m{text}\u001b[0m"
+
+
+def print_text(
+    text: str, color: Optional[str] = None, end: str = "", file: Optional[TextIO] = None
+) -> None:
+    """Print text with highlighting and no end characters."""
+    text_to_print = get_colored_text(text, color) if color else text
+    print(text_to_print, end=end, file=file)
+    if file:
+        file.flush()  # ensure all printed content are written to file
+
+class DifyAgentCallbackHandler(BaseModel):
    """Callback Handler that prints to std out."""
    color: Optional[str] = ''
    current_loop = 1
@@ -36,7 +56,7 @@ class DifyAgentCallbackHandler(BaseCallbackHandler, BaseModel):
        print_text("\n[on_tool_end]\n", color=self.color)
        print_text("Tool: " + tool_name + "\n", color=self.color)
        print_text("Inputs: " + str(tool_inputs) + "\n", color=self.color)
-        print_text("Outputs: " + str(tool_outputs) + "\n", color=self.color)
+        print_text("Outputs: " + str(tool_outputs)[:1000] + "\n", color=self.color)
        print_text("\n")

    def on_tool_error(
@@ -0,0 +1,5 @@
+from core.callback_handler.agent_tool_callback_handler import DifyAgentCallbackHandler
+
+
+class DifyWorkflowCallbackHandler(DifyAgentCallbackHandler):
+    """Callback Handler that prints to std out."""
@@ -12,6 +12,7 @@ from core.rag.datasource.entity.embedding import Embeddings
 from extensions.ext_database import db
 from extensions.ext_redis import redis_client
 from libs import helper
+from models.dataset import Embedding

 logger = logging.getLogger(__name__)

@@ -23,32 +24,62 @@ class CacheEmbedding(Embeddings):

    def embed_documents(self, texts: list[str]) -> list[list[float]]:
        """Embed search docs in batches of 10."""
-        text_embeddings = []
-        try:
-            model_type_instance = cast(TextEmbeddingModel, self._model_instance.model_type_instance)
-            model_schema = model_type_instance.get_model_schema(self._model_instance.model, self._model_instance.credentials)
-            max_chunks = model_schema.model_properties[ModelPropertyKey.MAX_CHUNKS] \
-                if model_schema and ModelPropertyKey.MAX_CHUNKS in model_schema.model_properties else 1
-            for i in range(0, len(texts), max_chunks):
-                batch_texts = texts[i:i + max_chunks]
+        # use doc embedding cache or store if not exists
+        text_embeddings = [None for _ in range(len(texts))]
+        embedding_queue_indices = []
+        for i, text in enumerate(texts):
+            hash = helper.generate_text_hash(text)
+            embedding = db.session.query(Embedding).filter_by(model_name=self._model_instance.model,
+                                                              hash=hash,
+                                                              provider_name=self._model_instance.provider).first()
+            if embedding:
+                text_embeddings[i] = embedding.get_embedding()
+            else:
+                embedding_queue_indices.append(i)
+        if embedding_queue_indices:
+            embedding_queue_texts = [texts[i] for i in embedding_queue_indices]
+            embedding_queue_embeddings = []
+            try:
+                model_type_instance = cast(TextEmbeddingModel, self._model_instance.model_type_instance)
+                model_schema = model_type_instance.get_model_schema(self._model_instance.model,
+                                                                    self._model_instance.credentials)
+                max_chunks = model_schema.model_properties[ModelPropertyKey.MAX_CHUNKS] \
+                    if model_schema and ModelPropertyKey.MAX_CHUNKS in model_schema.model_properties else 1
+                for i in range(0, len(embedding_queue_texts), max_chunks):
+                    batch_texts = embedding_queue_texts[i:i + max_chunks]

-                embedding_result = self._model_instance.invoke_text_embedding(
-                    texts=batch_texts,
-                    user=self._user
-                )
+                    embedding_result = self._model_instance.invoke_text_embedding(
+                        texts=batch_texts,
+                        user=self._user
+                    )

-                for vector in embedding_result.embeddings:
-                    try:
-                        normalized_embedding = (vector / np.linalg.norm(vector)).tolist()
-                        text_embeddings.append(normalized_embedding)
-                    except IntegrityError:
-                        db.session.rollback()
-                    except Exception as e:
-                        logging.exception('Failed to add embedding to redis')
-
-        except Exception as ex:
-            logger.error('Failed to embed documents: ', ex)
-            raise ex
+                    for vector in embedding_result.embeddings:
+                        try:
+                            normalized_embedding = (vector / np.linalg.norm(vector)).tolist()
+                            embedding_queue_embeddings.append(normalized_embedding)
+                        except IntegrityError:
+                            db.session.rollback()
+                        except Exception as e:
+                            logging.exception('Failed transform embedding: ', e)
+                cache_embeddings = []
+                try:
+                    for i, embedding in zip(embedding_queue_indices, embedding_queue_embeddings):
+                        text_embeddings[i] = embedding
+                        hash = helper.generate_text_hash(texts[i])
+                        if hash not in cache_embeddings:
+                            embedding_cache = Embedding(model_name=self._model_instance.model,
+                                                        hash=hash,
+                                                        provider_name=self._model_instance.provider)
+                            embedding_cache.set_embedding(embedding)
+                            db.session.add(embedding_cache)
+                            cache_embeddings.append(hash)
+                    db.session.commit()
+                except IntegrityError:
+                    db.session.rollback()
+            except Exception as ex:
+                db.session.rollback()
+                logger.error('Failed to embed documents: ', ex)
+                raise ex

        return text_embeddings

@@ -61,8 +92,6 @@ class CacheEmbedding(Embeddings):
        if embedding:
            redis_client.expire(embedding_cache_key, 600)
            return list(np.frombuffer(base64.b64decode(embedding), dtype="float"))
-
-
        try:
            embedding_result = self._model_instance.invoke_text_embedding(
                texts=[text],
@@ -1,5 +1,5 @@
 import enum
-import importlib.util
+import importlib
 import json
 import logging
 import os
@@ -47,7 +47,7 @@ class CodeExecutor:
        else:
            raise CodeExecutionException('Unsupported language')

-        runner = template_transformer.transform_caller(code, inputs)
+        runner, preload = template_transformer.transform_caller(code, inputs)
        url = URL(CODE_EXECUTION_ENDPOINT) / 'v1' / 'sandbox' / 'run'
        headers = {
            'X-Api-Key': CODE_EXECUTION_API_KEY
@@ -57,6 +57,7 @@ class CodeExecutor:
                        'nodejs' if language == 'javascript' else
                        'python3' if language == 'python3' else None,
            'code': runner,
+            'preload': preload
        }

        try:
@@ -64,11 +65,11 @@ class CodeExecutor:
            if response.status_code == 503:
                raise CodeExecutionException('Code execution service is unavailable')
            elif response.status_code != 200:
-                raise Exception('Failed to execute code')
+                raise Exception(f'Failed to execute code, got status code {response.status_code}, please check if the sandbox service is running')
        except CodeExecutionException as e:
            raise e
        except Exception as e:
-            raise CodeExecutionException('Failed to execute code')
+            raise CodeExecutionException('Failed to execute code, this is likely a network issue, please check if the sandbox service is running')
        
        try:
            response = response.json()
@@ -18,26 +18,27 @@ result = `<<RESULT>>${output}<<RESULT>>`
 console.log(result)
 """

+NODEJS_PRELOAD = """"""

 class NodeJsTemplateTransformer(TemplateTransformer):
    @classmethod
-    def transform_caller(cls, code: str, inputs: dict) -> str:
+    def transform_caller(cls, code: str, inputs: dict) -> tuple[str, str]:
        """
        Transform code to python runner
        :param code: code
        :param inputs: inputs
        :return:
        """
-        
+
        # transform inputs to json string
-        inputs_str = json.dumps(inputs, indent=4)
+        inputs_str = json.dumps(inputs, indent=4, ensure_ascii=False)

        # replace code and inputs
        runner = NODEJS_RUNNER.replace('{{code}}', code)
        runner = runner.replace('{{inputs}}', inputs_str)

-        return runner
-    
+        return runner, NODEJS_PRELOAD
+
    @classmethod
    def transform_response(cls, response: str) -> dict:
        """
@@ -20,9 +20,39 @@ print(result)

 """

+JINJA2_PRELOAD_TEMPLATE = """{% set fruits = ['Apple'] %}
+{{ 'a' }}
+{% for fruit in fruits %}
+    <li>{{ fruit }}</li>
+{% endfor %}
+{% if fruits|length > 1 %}
+1
+{% endif %}
+{% for i in range(5) %}
+    {% if i == 3 %}{{ i }}{% else %}{% endif %}
+{% endfor %}
+    {% for i in range(3) %}
+        {{ i + 1 }}
+    {% endfor %}
+{% macro say_hello() %}a{{ 'b' }}{% endmacro %}
+{{ s }}{{ say_hello() }}"""
+
+JINJA2_PRELOAD = f"""
+import jinja2
+
+def _jinja2_preload_():
+    # prepare jinja2 environment, load template and render before to avoid sandbox issue
+    template = jinja2.Template('''{JINJA2_PRELOAD_TEMPLATE}''')
+    template.render(s='a')
+
+if __name__ == '__main__':
+    _jinja2_preload_()
+
+"""
+
 class Jinja2TemplateTransformer(TemplateTransformer):
    @classmethod
-    def transform_caller(cls, code: str, inputs: dict) -> str:
+    def transform_caller(cls, code: str, inputs: dict) -> tuple[str, str]:
        """
        Transform code to python runner
        :param code: code
@@ -32,10 +62,10 @@ class Jinja2TemplateTransformer(TemplateTransformer):

        # transform jinja2 template to python code
        runner = PYTHON_RUNNER.replace('{{code}}', code)
-        runner = runner.replace('{{inputs}}', json.dumps(inputs, indent=4))
+        runner = runner.replace('{{inputs}}', json.dumps(inputs, indent=4, ensure_ascii=False))
+
+        return runner, JINJA2_PRELOAD

-        return runner
-    
    @classmethod
    def transform_response(cls, response: str) -> dict:
        """
@@ -51,4 +81,4 @@ class Jinja2TemplateTransformer(TemplateTransformer):

        return {
            'result': result
-        }
+        }
@@ -20,10 +20,12 @@ result = f'''<<RESULT>>
 print(result)
 """

+PYTHON_PRELOAD = """"""
+

 class PythonTemplateTransformer(TemplateTransformer):
    @classmethod
-    def transform_caller(cls, code: str, inputs: dict) -> str:
+    def transform_caller(cls, code: str, inputs: dict) -> tuple[str, str]:
        """
        Transform code to python runner
        :param code: code
@@ -32,13 +34,13 @@ class PythonTemplateTransformer(TemplateTransformer):
        """
        
        # transform inputs to json string
-        inputs_str = json.dumps(inputs, indent=4)
+        inputs_str = json.dumps(inputs, indent=4, ensure_ascii=False)

        # replace code and inputs
        runner = PYTHON_RUNNER.replace('{{code}}', code)
        runner = runner.replace('{{inputs}}', inputs_str)

-        return runner
+        return runner, PYTHON_PRELOAD
    
    @classmethod
    def transform_response(cls, response: str) -> dict:
@@ -4,12 +4,12 @@ from abc import ABC, abstractmethod
 class TemplateTransformer(ABC):
    @classmethod
    @abstractmethod
-    def transform_caller(cls, code: str, inputs: dict) -> str:
+    def transform_caller(cls, code: str, inputs: dict) -> tuple[str, str]:
        """
        Transform code to python runner
        :param code: code
        :param inputs: inputs
-        :return:
+        :return: runner, preload
        """
        pass
    
@@ -1,3 +1,4 @@
+import concurrent.futures
 import datetime
 import json
 import logging
@@ -18,6 +19,7 @@ from core.model_manager import ModelInstance, ModelManager
 from core.model_runtime.entities.model_entities import ModelType, PriceType
 from core.model_runtime.model_providers.__base.large_language_model import LargeLanguageModel
 from core.model_runtime.model_providers.__base.text_embedding_model import TextEmbeddingModel
+from core.rag.datasource.keyword.keyword_factory import Keyword
 from core.rag.extractor.entity.extract_setting import ExtractSetting
 from core.rag.index_processor.index_processor_base import BaseIndexProcessor
 from core.rag.index_processor.index_processor_factory import IndexProcessorFactory
@@ -650,17 +652,72 @@ class IndexingRunner:
        # chunk nodes by chunk size
        indexing_start_at = time.perf_counter()
        tokens = 0
-        chunk_size = 100
+        chunk_size = 10

        embedding_model_type_instance = None
        if embedding_model_instance:
            embedding_model_type_instance = embedding_model_instance.model_type_instance
            embedding_model_type_instance = cast(TextEmbeddingModel, embedding_model_type_instance)
+        # create keyword index
+        create_keyword_thread = threading.Thread(target=self._process_keyword_index,
+                                                 args=(current_app._get_current_object(),
+                                                       dataset.id, dataset_document.id, documents))
+        create_keyword_thread.start()
+        if dataset.indexing_technique == 'high_quality':
+            with concurrent.futures.ThreadPoolExecutor(max_workers=10) as executor:
+                futures = []
+                for i in range(0, len(documents), chunk_size):
+                    chunk_documents = documents[i:i + chunk_size]
+                    futures.append(executor.submit(self._process_chunk, current_app._get_current_object(), index_processor,
+                                                   chunk_documents, dataset,
+                                                   dataset_document, embedding_model_instance,
+                                                   embedding_model_type_instance))

-        for i in range(0, len(documents), chunk_size):
+                for future in futures:
+                    tokens += future.result()
+
+        create_keyword_thread.join()
+        indexing_end_at = time.perf_counter()
+
+        # update document status to completed
+        self._update_document_index_status(
+            document_id=dataset_document.id,
+            after_indexing_status="completed",
+            extra_update_params={
+                DatasetDocument.tokens: tokens,
+                DatasetDocument.completed_at: datetime.datetime.utcnow(),
+                DatasetDocument.indexing_latency: indexing_end_at - indexing_start_at,
+            }
+        )
+
+    def _process_keyword_index(self, flask_app, dataset_id, document_id, documents):
+        with flask_app.app_context():
+            dataset = Dataset.query.filter_by(id=dataset_id).first()
+            if not dataset:
+                raise ValueError("no dataset found")
+            keyword = Keyword(dataset)
+            keyword.create(documents)
+            if dataset.indexing_technique != 'high_quality':
+                document_ids = [document.metadata['doc_id'] for document in documents]
+                db.session.query(DocumentSegment).filter(
+                    DocumentSegment.document_id == document_id,
+                    DocumentSegment.index_node_id.in_(document_ids),
+                    DocumentSegment.status == "indexing"
+                ).update({
+                    DocumentSegment.status: "completed",
+                    DocumentSegment.enabled: True,
+                    DocumentSegment.completed_at: datetime.datetime.utcnow()
+                })
+
+                db.session.commit()
+
+    def _process_chunk(self, flask_app, index_processor, chunk_documents, dataset, dataset_document,
+                       embedding_model_instance, embedding_model_type_instance):
+        with flask_app.app_context():
            # check document is paused
            self._check_document_paused_status(dataset_document.id)
-            chunk_documents = documents[i:i + chunk_size]
+
+            tokens = 0
            if dataset.indexing_technique == 'high_quality' or embedding_model_type_instance:
                tokens += sum(
                    embedding_model_type_instance.get_num_tokens(
@@ -670,9 +727,9 @@ class IndexingRunner:
                    )
                    for document in chunk_documents
                )
+
            # load index
-            index_processor.load(dataset, chunk_documents)
-            db.session.add(dataset)
+            index_processor.load(dataset, chunk_documents, with_keywords=False)

            document_ids = [document.metadata['doc_id'] for document in chunk_documents]
            db.session.query(DocumentSegment).filter(
@@ -687,18 +744,7 @@ class IndexingRunner:

            db.session.commit()

-        indexing_end_at = time.perf_counter()
-
-        # update document status to completed
-        self._update_document_index_status(
-            document_id=dataset_document.id,
-            after_indexing_status="completed",
-            extra_update_params={
-                DatasetDocument.tokens: tokens,
-                DatasetDocument.completed_at: datetime.datetime.utcnow(),
-                DatasetDocument.indexing_latency: indexing_end_at - indexing_start_at,
-            }
-        )
+            return tokens

    def _check_document_paused_status(self, document_id: str):
        indexing_cache_key = 'document_{}_is_paused'.format(document_id)
@@ -166,6 +166,7 @@ class LLMGenerator:
        response = model_instance.invoke_llm(
            prompt_messages=prompt_messages,
            model_parameters={
+                'temperature': 0.01,
                "max_tokens": 2000
            },
            stream=False
@@ -69,13 +69,15 @@ SUGGESTED_QUESTIONS_AFTER_ANSWER_INSTRUCTION_PROMPT = (
 )

 GENERATOR_QA_PROMPT = (
-    'The user will send a long text. Please think step by step.'
+    '<Task> The user will send a long text. Generate a Question and Answer pairs only using the knowledge in the long text. Please think step by step.'
    'Step 1: Understand and summarize the main content of this text.\n'
    'Step 2: What key information or concepts are mentioned in this text?\n'
    'Step 3: Decompose or combine multiple pieces of information and concepts.\n'
-    'Step 4: Generate 20 questions and answers based on these key information and concepts.'
-    'The questions should be clear and detailed, and the answers should be detailed and complete.\n'
-    "Answer MUST according to the the language:{language} and in the following format: Q1:\nA1:\nQ2:\nA2:...\n"
+    'Step 4: Generate questions and answers based on these key information and concepts.\n'
+    '<Constraints> The questions should be clear and detailed, and the answers should be detailed and complete. '
+    'You must answer in {language}, in a style that is clear and detailed in {language}. No language other than {language} should be used. \n'
+    '<Format> Use the following format: Q1:\nA1:\nQ2:\nA2:...\n'
+    '<QA Pairs>'
 )

 RULE_CONFIG_GENERATE_TEMPLATE = """Given MY INTENDED AUDIENCES and HOPING TO SOLVE using a language model, please select \
@@ -3,6 +3,7 @@ from core.file.message_file_parser import MessageFileParser
 from core.model_manager import ModelInstance
 from core.model_runtime.entities.message_entities import (
    AssistantPromptMessage,
+    ImagePromptMessageContent,
    PromptMessage,
    PromptMessageRole,
    TextPromptMessageContent,
@@ -124,7 +125,17 @@ class TokenBufferMemory:
            else:
                continue

-            message = f"{role}: {m.content}"
-            string_messages.append(message)
+            if isinstance(m.content, list):
+                inner_msg = ""
+                for content in m.content:
+                    if isinstance(content, TextPromptMessageContent):
+                        inner_msg += f"{content.data}\n"
+                    elif isinstance(content, ImagePromptMessageContent):
+                        inner_msg += "[image]\n"
+
+                string_messages.append(f"{role}: {inner_msg.strip()}")
+            else:
+                message = f"{role}: {m.content}"
+                string_messages.append(message)

        return "\n".join(string_messages)
@@ -147,7 +147,7 @@

 - `input` (float) Input price, i.e., Prompt price
 - `output` (float) Output price, i.e., returned content price
- `unit` (float) Pricing unit, e.g., per 100K price is `0.000001`
+- `unit` (float) Pricing unit, e.g., if the price is meausred in 1M tokens, the corresponding token amount for the unit price is `0.000001`.
 - `currency` (string) Currency unit

 ### ProviderCredentialSchema
@@ -149,7 +149,7 @@

 - `input` (float) 输入单价，即 Prompt 单价
 - `output` (float) 输出单价，即返回内容单价
- `unit` (float) 价格单位，如：每 100K 的单价为 `0.000001`
+- `unit` (float) 价格单位，如以 1M tokens 计价，则单价对应的单位 token 数为 `0.000001`
 - `currency` (string) 货币单位

 ### ProviderCredentialSchema
@@ -73,8 +73,8 @@ PARAMETER_RULE_TEMPLATE: dict[DefaultParameterName, dict] = {
        },
        'type': 'int',
        'help': {
-            'en_US': 'The maximum number of tokens to generate. Requests can use up to 2048 tokens shared between prompt and completion.',
-            'zh_Hans': '要生成的标记的最大数量。请求可以使用最多2048个标记，这些标记在提示和完成之间共享。',
+            'en_US': 'Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter.',
+            'zh_Hans': '指定生成结果长度的上限。如果生成结果截断，可以调大该参数。',
        },
        'required': False,
        'default': 64,
@@ -1,4 +1,3 @@
-import importlib
 import os
 from abc import ABC, abstractmethod

@@ -7,6 +6,7 @@ import yaml
 from core.model_runtime.entities.model_entities import AIModelEntity, ModelType
 from core.model_runtime.entities.provider_entities import ProviderEntity
 from core.model_runtime.model_providers.__base.ai_model import AIModel
+from core.utils.module_import_helper import get_subclasses_from_module, import_module_from_source


 class ModelProvider(ABC):
@@ -104,17 +104,10 @@ class ModelProvider(ABC):

        # Dynamic loading {model_type_name}.py file and find the subclass of AIModel
        parent_module = '.'.join(self.__class__.__module__.split('.')[:-1])
-        spec = importlib.util.spec_from_file_location(f"{parent_module}.{model_type_name}.{model_type_name}", model_type_py_path)
-        mod = importlib.util.module_from_spec(spec)
-        spec.loader.exec_module(mod)
-
-        model_class = None
-        for name, obj in vars(mod).items():
-            if (isinstance(obj, type) and issubclass(obj, AIModel) and not obj.__abstractmethods__
-                    and obj != AIModel and obj.__module__ == mod.__name__):
-                model_class = obj
-                break
-
+        mod = import_module_from_source(
+            f'{parent_module}.{model_type_name}.{model_type_name}', model_type_py_path)
+        model_class = next(filter(lambda x: x.__module__ == mod.__name__ and not x.__abstractmethods__,
+                                  get_subclasses_from_module(mod, AIModel)), None)
        if not model_class:
            raise Exception(f'Missing AIModel Class for model type {model_type} in {model_type_py_path}')

@@ -6,6 +6,7 @@
 - cohere
 - bedrock
 - togetherai
+- openrouter
 - ollama
 - mistralai
 - groq
@@ -5,6 +5,7 @@ model_type: llm
 features:
  - agent-thought
  - vision
+  - tool-call
 model_properties:
  mode: chat
  context_size: 200000
@@ -5,6 +5,7 @@ model_type: llm
 features:
  - agent-thought
  - vision
+  - tool-call
 model_properties:
  mode: chat
  context_size: 200000
@@ -5,6 +5,7 @@ model_type: llm
 features:
  - agent-thought
  - vision
+  - tool-call
 model_properties:
  mode: chat
  context_size: 200000
@@ -1,4 +1,5 @@
 import base64
+import json
 import mimetypes
 from collections.abc import Generator
 from typing import Optional, Union, cast
@@ -15,6 +16,7 @@ from anthropic.types import (
    MessageStreamEvent,
    completion_create_params,
 )
+from anthropic.types.beta.tools import ToolsBetaMessage
 from httpx import Timeout

 from core.model_runtime.callbacks.base_callback import Callback
@@ -27,6 +29,7 @@ from core.model_runtime.entities.message_entities import (
    PromptMessageTool,
    SystemPromptMessage,
    TextPromptMessageContent,
+    ToolPromptMessage,
    UserPromptMessage,
 )
 from core.model_runtime.errors.invoke import (
@@ -70,10 +73,11 @@ class AnthropicLargeLanguageModel(LargeLanguageModel):
        :return: full response or stream response chunk generator result
        """
        # invoke model
-        return self._chat_generate(model, credentials, prompt_messages, model_parameters, stop, stream, user)
+        return self._chat_generate(model, credentials, prompt_messages, model_parameters, tools, stop, stream, user)

    def _chat_generate(self, model: str, credentials: dict,
-                       prompt_messages: list[PromptMessage], model_parameters: dict, stop: Optional[list[str]] = None,
+                       prompt_messages: list[PromptMessage], model_parameters: dict, 
+                       tools: Optional[list[PromptMessageTool]] = None, stop: Optional[list[str]] = None,
                       stream: bool = True, user: Optional[str] = None) -> Union[LLMResult, Generator]:
        """
        Invoke llm chat model
@@ -109,14 +113,26 @@ class AnthropicLargeLanguageModel(LargeLanguageModel):
        if system:
            extra_model_kwargs['system'] = system

-        # chat model
-        response = client.messages.create(
-            model=model,
-            messages=prompt_message_dicts,
-            stream=stream,
-            **model_parameters,
-            **extra_model_kwargs
-        )
+        if tools:
+            extra_model_kwargs['tools'] = [
+                self._transform_tool_prompt(tool) for tool in tools
+            ]
+            response = client.beta.tools.messages.create(
+                model=model,
+                messages=prompt_message_dicts,
+                stream=stream,
+                **model_parameters,
+                **extra_model_kwargs
+            )
+        else:
+            # chat model
+            response = client.messages.create(
+                model=model,
+                messages=prompt_message_dicts,
+                stream=stream,
+                **model_parameters,
+                **extra_model_kwargs
+            )

        if stream:
            return self._handle_chat_generate_stream_response(model, credentials, response, prompt_messages)
@@ -148,6 +164,13 @@ class AnthropicLargeLanguageModel(LargeLanguageModel):

        return self._invoke(model, credentials, prompt_messages, model_parameters, tools, stop, stream, user)

+    def _transform_tool_prompt(self, tool: PromptMessageTool) -> dict:
+        return {
+            'name': tool.name,
+            'description': tool.description,
+            'input_schema': tool.parameters
+        }
+
    def _transform_chat_json_prompts(self, model: str, credentials: dict,
                                     prompt_messages: list[PromptMessage], model_parameters: dict,
                                     tools: list[PromptMessageTool] | None = None, stop: list[str] | None = None,
@@ -193,7 +216,18 @@ class AnthropicLargeLanguageModel(LargeLanguageModel):
        prompt = self._convert_messages_to_prompt_anthropic(prompt_messages)

        client = Anthropic(api_key="")
-        return client.count_tokens(prompt)
+        tokens = client.count_tokens(prompt)
+
+        tool_call_inner_prompts_tokens_map = {
+            'claude-3-opus-20240229': 395,
+            'claude-3-haiku-20240307': 264,
+            'claude-3-sonnet-20240229': 159
+        }
+
+        if model in tool_call_inner_prompts_tokens_map and tools:
+            tokens += tool_call_inner_prompts_tokens_map[model]
+
+        return tokens

    def validate_credentials(self, model: str, credentials: dict) -> None:
        """
@@ -219,7 +253,7 @@ class AnthropicLargeLanguageModel(LargeLanguageModel):
        except Exception as ex:
            raise CredentialsValidateFailedError(str(ex))

-    def _handle_chat_generate_response(self, model: str, credentials: dict, response: Message,
+    def _handle_chat_generate_response(self, model: str, credentials: dict, response: Union[Message, ToolsBetaMessage],
                                       prompt_messages: list[PromptMessage]) -> LLMResult:
        """
        Handle llm chat response
@@ -232,9 +266,24 @@ class AnthropicLargeLanguageModel(LargeLanguageModel):
        """
        # transform assistant message to prompt message
        assistant_prompt_message = AssistantPromptMessage(
-            content=response.content[0].text
+            content='',
+            tool_calls=[]
        )

+        for content in response.content:
+            if content.type == 'text':
+                assistant_prompt_message.content += content.text
+            elif content.type == 'tool_use':
+                tool_call = AssistantPromptMessage.ToolCall(
+                    id=content.id,
+                    type='function',
+                    function=AssistantPromptMessage.ToolCall.ToolCallFunction(
+                        name=content.name,
+                        arguments=json.dumps(content.input)
+                    )
+                )
+                assistant_prompt_message.tool_calls.append(tool_call)
+
        # calculate num tokens
        if response.usage:
            # transform usage
@@ -356,69 +405,90 @@ class AnthropicLargeLanguageModel(LargeLanguageModel):
        prompt_message_dicts = []
        for message in prompt_messages:
            if not isinstance(message, SystemPromptMessage):
-                prompt_message_dicts.append(self._convert_prompt_message_to_dict(message))
+                if isinstance(message, UserPromptMessage):
+                    message = cast(UserPromptMessage, message)
+                    if isinstance(message.content, str):
+                        message_dict = {"role": "user", "content": message.content}
+                        prompt_message_dicts.append(message_dict)
+                    else:
+                        sub_messages = []
+                        for message_content in message.content:
+                            if message_content.type == PromptMessageContentType.TEXT:
+                                message_content = cast(TextPromptMessageContent, message_content)
+                                sub_message_dict = {
+                                    "type": "text",
+                                    "text": message_content.data
+                                }
+                                sub_messages.append(sub_message_dict)
+                            elif message_content.type == PromptMessageContentType.IMAGE:
+                                message_content = cast(ImagePromptMessageContent, message_content)
+                                if not message_content.data.startswith("data:"):
+                                    # fetch image data from url
+                                    try:
+                                        image_content = requests.get(message_content.data).content
+                                        mime_type, _ = mimetypes.guess_type(message_content.data)
+                                        base64_data = base64.b64encode(image_content).decode('utf-8')
+                                    except Exception as ex:
+                                        raise ValueError(f"Failed to fetch image data from url {message_content.data}, {ex}")
+                                else:
+                                    data_split = message_content.data.split(";base64,")
+                                    mime_type = data_split[0].replace("data:", "")
+                                    base64_data = data_split[1]
+
+                                if mime_type not in ["image/jpeg", "image/png", "image/gif", "image/webp"]:
+                                    raise ValueError(f"Unsupported image type {mime_type}, "
+                                                    f"only support image/jpeg, image/png, image/gif, and image/webp")
+
+                                sub_message_dict = {
+                                    "type": "image",
+                                    "source": {
+                                        "type": "base64",
+                                        "media_type": mime_type,
+                                        "data": base64_data
+                                    }
+                                }
+                                sub_messages.append(sub_message_dict)
+                        prompt_message_dicts.append({"role": "user", "content": sub_messages})
+                elif isinstance(message, AssistantPromptMessage):
+                    message = cast(AssistantPromptMessage, message)
+                    content = []
+                    if message.tool_calls:
+                        for tool_call in message.tool_calls:
+                            content.append({
+                                "type": "tool_use",
+                                "id": tool_call.id,
+                                "name": tool_call.function.name,
+                                "input": json.loads(tool_call.function.arguments)
+                            })
+                    if message.content:
+                        content.append({
+                            "type": "text",
+                            "text": message.content
+                        })
+                    
+                    if prompt_message_dicts[-1]["role"] == "assistant":
+                        prompt_message_dicts[-1]["content"].extend(content)
+                    else:
+                        prompt_message_dicts.append({
+                            "role": "assistant",
+                            "content": content
+                        })
+                elif isinstance(message, ToolPromptMessage):
+                    message = cast(ToolPromptMessage, message)
+                    message_dict = {
+                        "role": "user",
+                        "content": [{
+                            "type": "tool_result",
+                            "tool_use_id": message.tool_call_id,
+                            "content": message.content
+                        }]
+                    }
+                    prompt_message_dicts.append(message_dict)
+                else:
+                    raise ValueError(f"Got unknown type {message}")

        return system, prompt_message_dicts

-    def _convert_prompt_message_to_dict(self, message: PromptMessage) -> dict:
-        """
-        Convert PromptMessage to dict
-        """
-        if isinstance(message, UserPromptMessage):
-            message = cast(UserPromptMessage, message)
-            if isinstance(message.content, str):
-                message_dict = {"role": "user", "content": message.content}
-            else:
-                sub_messages = []
-                for message_content in message.content:
-                    if message_content.type == PromptMessageContentType.TEXT:
-                        message_content = cast(TextPromptMessageContent, message_content)
-                        sub_message_dict = {
-                            "type": "text",
-                            "text": message_content.data
-                        }
-                        sub_messages.append(sub_message_dict)
-                    elif message_content.type == PromptMessageContentType.IMAGE:
-                        message_content = cast(ImagePromptMessageContent, message_content)
-                        if not message_content.data.startswith("data:"):
-                            # fetch image data from url
-                            try:
-                                image_content = requests.get(message_content.data).content
-                                mime_type, _ = mimetypes.guess_type(message_content.data)
-                                base64_data = base64.b64encode(image_content).decode('utf-8')
-                            except Exception as ex:
-                                raise ValueError(f"Failed to fetch image data from url {message_content.data}, {ex}")
-                        else:
-                            data_split = message_content.data.split(";base64,")
-                            mime_type = data_split[0].replace("data:", "")
-                            base64_data = data_split[1]
-
-                        if mime_type not in ["image/jpeg", "image/png", "image/gif", "image/webp"]:
-                            raise ValueError(f"Unsupported image type {mime_type}, "
-                                             f"only support image/jpeg, image/png, image/gif, and image/webp")
-
-                        sub_message_dict = {
-                            "type": "image",
-                            "source": {
-                                "type": "base64",
-                                "media_type": mime_type,
-                                "data": base64_data
-                            }
-                        }
-                        sub_messages.append(sub_message_dict)
-
-                message_dict = {"role": "user", "content": sub_messages}
-        elif isinstance(message, AssistantPromptMessage):
-            message = cast(AssistantPromptMessage, message)
-            message_dict = {"role": "assistant", "content": message.content}
-        elif isinstance(message, SystemPromptMessage):
-            message = cast(SystemPromptMessage, message)
-            message_dict = {"role": "system", "content": message.content}
-        else:
-            raise ValueError(f"Got unknown type {message}")
-
-        return message_dict
-
    def _convert_one_message_to_text(self, message: PromptMessage) -> str:
        """
        Convert a single message to a string.
@@ -453,6 +523,8 @@ class AnthropicLargeLanguageModel(LargeLanguageModel):
                        message_text += f"{ai_prompt} [IMAGE]"
        elif isinstance(message, SystemPromptMessage):
            message_text = content
+        elif isinstance(message, ToolPromptMessage):
+            message_text = f"{human_prompt} {message.content}"
        else:
            raise ValueError(f"Got unknown type {message}")

@@ -17,9 +17,11 @@ class BedrockProvider(ModelProvider):
        """
        try:
            model_instance = self.get_model_instance(ModelType.LLM)
-            bedrock_validate_model_name = credentials.get('model_for_validation', 'amazon.titan-text-lite-v1')
+
+            # Use `amazon.titan-text-lite-v1` model by default for validating credentials
+            model_for_validation = credentials.get('model_for_validation', 'amazon.titan-text-lite-v1')
            model_instance.validate_credentials(
-                model=bedrock_validate_model_name,
+                model=model_for_validation,
                credentials=credentials
            )
        except CredentialsValidateFailedError as ex:
@@ -1,33 +1,50 @@
 model: anthropic.claude-instant-v1
 label:
-  en_US: Claude Instant V1
+  en_US: Claude Instant 1
 model_type: llm
 model_properties:
  mode: chat
  context_size: 100000
 parameter_rules:
-  - name: temperature
-    use_template: temperature
-  - name: topP
-    use_template: top_p
-  - name: topK
-    label:
-      zh_Hans: 取样数量
-      en_US: Top K
-    type: int
-    help:
-      zh_Hans: 仅从每个后续标记的前 K 个选项中采样。
-      en_US: Only sample from the top K options for each subsequent token.
-    required: false
-    default: 250
-    min: 0
-    max: 500
-  - name: max_tokens_to_sample
+  - name: max_tokens
    use_template: max_tokens
    required: true
+    type: int
    default: 4096
    min: 1
    max: 4096
+    help:
+      zh_Hans: 停止前生成的最大令牌数。请注意，Anthropic Claude 模型可能会在达到 max_tokens 的值之前停止生成令牌。不同的 Anthropic Claude 模型对此参数具有不同的最大值。
+      en_US: The maximum number of tokens to generate before stopping. Note that Anthropic Claude models might stop generating tokens before reaching the value of max_tokens. Different Anthropic Claude models have different maximum values for this parameter.
+  - name: temperature
+    use_template: temperature
+    required: false
+    type: float
+    default: 1
+    min: 0.0
+    max: 1.0
+    help:
+      zh_Hans: 生成内容的随机性。
+      en_US: The amount of randomness injected into the response.
+  - name: top_p
+    required: false
+    type: float
+    default: 0.999
+    min: 0.000
+    max: 1.000
+    help:
+      zh_Hans: 在核采样中，Anthropic Claude 按概率递减顺序计算每个后续标记的所有选项的累积分布，并在达到 top_p 指定的特定概率时将其切断。您应该更改温度或top_p，但不能同时更改两者。
+      en_US: In nucleus sampling, Anthropic Claude computes the cumulative distribution over all the options for each subsequent token in decreasing probability order and cuts it off once it reaches a particular probability specified by top_p. You should alter either temperature or top_p, but not both.
+  - name: top_k
+    required: false
+    type: int
+    default: 0
+    min: 0
+    # tip docs from aws has error, max value is 500
+    max: 500
+    help:
+      zh_Hans: 对于每个后续标记，仅从前 K 个选项中进行采样。使用 top_k 删除长尾低概率响应。
+      en_US: Only sample from the top K options for each subsequent token. Use top_k to remove long tail low probability responses.
 pricing:
  input: '0.0008'
  output: '0.0024'
@@ -1,33 +1,50 @@
 model: anthropic.claude-v1
 label:
-  en_US: Claude V1
+  en_US: Claude 1
 model_type: llm
 model_properties:
  mode: chat
  context_size: 100000
 parameter_rules:
-  - name: temperature
-    use_template: temperature
-  - name: top_p
-    use_template: top_p
-  - name: top_k
-    label:
-      zh_Hans: 取样数量
-      en_US: Top K
-    type: int
-    help:
-      zh_Hans: 仅从每个后续标记的前 K 个选项中采样。
-      en_US: Only sample from the top K options for each subsequent token.
-    required: false
-    default: 250
-    min: 0
-    max: 500
-  - name: max_tokens_to_sample
+  - name: max_tokens
    use_template: max_tokens
    required: true
+    type: int
    default: 4096
    min: 1
    max: 4096
+    help:
+      zh_Hans: 停止前生成的最大令牌数。请注意，Anthropic Claude 模型可能会在达到 max_tokens 的值之前停止生成令牌。不同的 Anthropic Claude 模型对此参数具有不同的最大值。
+      en_US: The maximum number of tokens to generate before stopping. Note that Anthropic Claude models might stop generating tokens before reaching the value of max_tokens. Different Anthropic Claude models have different maximum values for this parameter.
+  - name: temperature
+    use_template: temperature
+    required: false
+    type: float
+    default: 1
+    min: 0.0
+    max: 1.0
+    help:
+      zh_Hans: 生成内容的随机性。
+      en_US: The amount of randomness injected into the response.
+  - name: top_p
+    required: false
+    type: float
+    default: 0.999
+    min: 0.000
+    max: 1.000
+    help:
+      zh_Hans: 在核采样中，Anthropic Claude 按概率递减顺序计算每个后续标记的所有选项的累积分布，并在达到 top_p 指定的特定概率时将其切断。您应该更改温度或top_p，但不能同时更改两者。
+      en_US: In nucleus sampling, Anthropic Claude computes the cumulative distribution over all the options for each subsequent token in decreasing probability order and cuts it off once it reaches a particular probability specified by top_p. You should alter either temperature or top_p, but not both.
+  - name: top_k
+    required: false
+    type: int
+    default: 0
+    min: 0
+    # tip docs from aws has error, max value is 500
+    max: 500
+    help:
+      zh_Hans: 对于每个后续标记，仅从前 K 个选项中进行采样。使用 top_k 删除长尾低概率响应。
+      en_US: Only sample from the top K options for each subsequent token. Use top_k to remove long tail low probability responses.
 pricing:
  input: '0.008'
  output: '0.024'
@@ -1,33 +1,50 @@
 model: anthropic.claude-v2:1
 label:
-  en_US: Claude V2.1
+  en_US: Claude 2.1
 model_type: llm
 model_properties:
  mode: chat
  context_size: 200000
 parameter_rules:
-  - name: temperature
-    use_template: temperature
-  - name: top_p
-    use_template: top_p
-  - name: top_k
-    label:
-      zh_Hans: 取样数量
-      en_US: Top K
-    type: int
-    help:
-      zh_Hans: 仅从每个后续标记的前 K 个选项中采样。
-      en_US: Only sample from the top K options for each subsequent token.
-    required: false
-    default: 250
-    min: 0
-    max: 500
-  - name: max_tokens_to_sample
+  - name: max_tokens
    use_template: max_tokens
    required: true
+    type: int
    default: 4096
    min: 1
    max: 4096
+    help:
+      zh_Hans: 停止前生成的最大令牌数。请注意，Anthropic Claude 模型可能会在达到 max_tokens 的值之前停止生成令牌。不同的 Anthropic Claude 模型对此参数具有不同的最大值。
+      en_US: The maximum number of tokens to generate before stopping. Note that Anthropic Claude models might stop generating tokens before reaching the value of max_tokens. Different Anthropic Claude models have different maximum values for this parameter.
+  - name: temperature
+    use_template: temperature
+    required: false
+    type: float
+    default: 1
+    min: 0.0
+    max: 1.0
+    help:
+      zh_Hans: 生成内容的随机性。
+      en_US: The amount of randomness injected into the response.
+  - name: top_p
+    required: false
+    type: float
+    default: 0.999
+    min: 0.000
+    max: 1.000
+    help:
+      zh_Hans: 在核采样中，Anthropic Claude 按概率递减顺序计算每个后续标记的所有选项的累积分布，并在达到 top_p 指定的特定概率时将其切断。您应该更改温度或top_p，但不能同时更改两者。
+      en_US: In nucleus sampling, Anthropic Claude computes the cumulative distribution over all the options for each subsequent token in decreasing probability order and cuts it off once it reaches a particular probability specified by top_p. You should alter either temperature or top_p, but not both.
+  - name: top_k
+    required: false
+    type: int
+    default: 0
+    min: 0
+    # tip docs from aws has error, max value is 500
+    max: 500
+    help:
+      zh_Hans: 对于每个后续标记，仅从前 K 个选项中进行采样。使用 top_k 删除长尾低概率响应。
+      en_US: Only sample from the top K options for each subsequent token. Use top_k to remove long tail low probability responses.
 pricing:
  input: '0.008'
  output: '0.024'
@@ -1,33 +1,50 @@
 model: anthropic.claude-v2
 label:
-  en_US: Claude V2
+  en_US: Claude 2
 model_type: llm
 model_properties:
  mode: chat
  context_size: 100000
 parameter_rules:
-  - name: temperature
-    use_template: temperature
-  - name: top_p
-    use_template: top_p
-  - name: top_k
-    label:
-      zh_Hans: 取样数量
-      en_US: Top K
-    type: int
-    help:
-      zh_Hans: 仅从每个后续标记的前 K 个选项中采样。
-      en_US: Only sample from the top K options for each subsequent token.
-    required: false
-    default: 250
-    min: 0
-    max: 500
-  - name: max_tokens_to_sample
+  - name: max_tokens
    use_template: max_tokens
    required: true
+    type: int
    default: 4096
    min: 1
    max: 4096
+    help:
+      zh_Hans: 停止前生成的最大令牌数。请注意，Anthropic Claude 模型可能会在达到 max_tokens 的值之前停止生成令牌。不同的 Anthropic Claude 模型对此参数具有不同的最大值。
+      en_US: The maximum number of tokens to generate before stopping. Note that Anthropic Claude models might stop generating tokens before reaching the value of max_tokens. Different Anthropic Claude models have different maximum values for this parameter.
+  - name: temperature
+    use_template: temperature
+    required: false
+    type: float
+    default: 1
+    min: 0.0
+    max: 1.0
+    help:
+      zh_Hans: 生成内容的随机性。
+      en_US: The amount of randomness injected into the response.
+  - name: top_p
+    required: false
+    type: float
+    default: 0.999
+    min: 0.000
+    max: 1.000
+    help:
+      zh_Hans: 在核采样中，Anthropic Claude 按概率递减顺序计算每个后续标记的所有选项的累积分布，并在达到 top_p 指定的特定概率时将其切断。您应该更改温度或top_p，但不能同时更改两者。
+      en_US: In nucleus sampling, Anthropic Claude computes the cumulative distribution over all the options for each subsequent token in decreasing probability order and cuts it off once it reaches a particular probability specified by top_p. You should alter either temperature or top_p, but not both.
+  - name: top_k
+    required: false
+    type: int
+    default: 0
+    min: 0
+    # tip docs from aws has error, max value is 500
+    max: 500
+    help:
+      zh_Hans: 对于每个后续标记，仅从前 K 个选项中进行采样。使用 top_k 删除长尾低概率响应。
+      en_US: Only sample from the top K options for each subsequent token. Use top_k to remove long tail low probability responses.
 pricing:
  input: '0.008'
  output: '0.024'
@@ -72,16 +72,16 @@ class BedrockLargeLanguageModel(LargeLanguageModel):
        :return: full response or stream response chunk generator result
        """

-        # invoke claude 3 models via anthropic official SDK
-        if "anthropic.claude-3" in model:
-            return self._invoke_claude3(model, credentials, prompt_messages, model_parameters, stop, stream, user)
-        # invoke model
+        # invoke anthropic models via anthropic official SDK
+        if "anthropic" in model:
+            return self._generate_anthropic(model, credentials, prompt_messages, model_parameters, stop, stream, user)
+        # invoke other models via boto3 client
        return self._generate(model, credentials, prompt_messages, model_parameters, stop, stream, user)

-    def _invoke_claude3(self, model: str, credentials: dict, prompt_messages: list[PromptMessage], model_parameters: dict,
+    def _generate_anthropic(self, model: str, credentials: dict, prompt_messages: list[PromptMessage], model_parameters: dict,
                stop: Optional[list[str]] = None, stream: bool = True, user: Optional[str] = None) -> Union[LLMResult, Generator]:
        """
-        Invoke Claude3 large language model
+        Invoke Anthropic large language model

        :param model: model name
        :param credentials: model credentials
@@ -114,7 +114,7 @@ class BedrockLargeLanguageModel(LargeLanguageModel):
            # ref: https://github.com/anthropics/anthropic-sdk-python/blob/e84645b07ca5267066700a104b4d8d6a8da1383d/src/anthropic/resources/messages.py#L465
            # extra_model_kwargs['metadata'] = message_create_params.Metadata(user_id=user)

-        system, prompt_message_dicts = self._convert_claude3_prompt_messages(prompt_messages)
+        system, prompt_message_dicts = self._convert_claude_prompt_messages(prompt_messages)

        if system:
            extra_model_kwargs['system'] = system
@@ -128,11 +128,11 @@ class BedrockLargeLanguageModel(LargeLanguageModel):
        )

        if stream:
-            return self._handle_claude3_stream_response(model, credentials, response, prompt_messages)
+            return self._handle_claude_stream_response(model, credentials, response, prompt_messages)

-        return self._handle_claude3_response(model, credentials, response, prompt_messages)
+        return self._handle_claude_response(model, credentials, response, prompt_messages)

-    def _handle_claude3_response(self, model: str, credentials: dict, response: Message,
+    def _handle_claude_response(self, model: str, credentials: dict, response: Message,
                                prompt_messages: list[PromptMessage]) -> LLMResult:
        """
        Handle llm chat response
@@ -172,7 +172,7 @@ class BedrockLargeLanguageModel(LargeLanguageModel):

        return response

-    def _handle_claude3_stream_response(self, model: str, credentials: dict, response: Stream[MessageStreamEvent],
+    def _handle_claude_stream_response(self, model: str, credentials: dict, response: Stream[MessageStreamEvent],
                                        prompt_messages: list[PromptMessage], ) -> Generator:
        """
        Handle llm chat stream response
@@ -231,7 +231,7 @@ class BedrockLargeLanguageModel(LargeLanguageModel):
        except Exception as ex:
            raise InvokeError(str(ex))

-    def _calc_claude3_response_usage(self, model: str, credentials: dict, prompt_tokens: int, completion_tokens: int) -> LLMUsage:
+    def _calc_claude_response_usage(self, model: str, credentials: dict, prompt_tokens: int, completion_tokens: int) -> LLMUsage:
        """
        Calculate response usage

@@ -275,7 +275,7 @@ class BedrockLargeLanguageModel(LargeLanguageModel):

        return usage

-    def _convert_claude3_prompt_messages(self, prompt_messages: list[PromptMessage]) -> tuple[str, list[dict]]:
+    def _convert_claude_prompt_messages(self, prompt_messages: list[PromptMessage]) -> tuple[str, list[dict]]:
        """
        Convert prompt messages to dict list and system
        """
@@ -295,11 +295,11 @@ class BedrockLargeLanguageModel(LargeLanguageModel):
        prompt_message_dicts = []
        for message in prompt_messages:
            if not isinstance(message, SystemPromptMessage):
-                prompt_message_dicts.append(self._convert_claude3_prompt_message_to_dict(message))
+                prompt_message_dicts.append(self._convert_claude_prompt_message_to_dict(message))

        return system, prompt_message_dicts

-    def _convert_claude3_prompt_message_to_dict(self, message: PromptMessage) -> dict:
+    def _convert_claude_prompt_message_to_dict(self, message: PromptMessage) -> dict:
        """
        Convert PromptMessage to dict
        """
@@ -402,25 +402,25 @@ class BedrockLargeLanguageModel(LargeLanguageModel):
        :param credentials: model credentials
        :return:
        """
-
-        if "anthropic.claude-3" in model:
-            try:
-                self._invoke_claude3(model=model,
-                                        credentials=credentials,
-                                        prompt_messages=[{"role": "user", "content": "ping"}],
-                                        model_parameters={},
-                                        stop=None,
-                                        stream=False)
-
-            except Exception as ex:
-                raise CredentialsValidateFailedError(str(ex))
-
+        required_params = {}
+        if "anthropic" in model:
+            required_params = {
+                "max_tokens": 32,
+            }
+        elif "ai21" in model:
+            # ValidationException: Malformed input request: #/temperature: expected type: Number, found: Null#/maxTokens: expected type: Integer, found: Null#/topP: expected type: Number, found: Null, please reformat your input and try again.
+            required_params = {
+                "temperature": 0.7,
+                "topP": 0.9,
+                "maxTokens": 32,
+            }
+            
        try:
            ping_message = UserPromptMessage(content="ping")
-            self._generate(model=model,
+            self._invoke(model=model,
                           credentials=credentials,
                           prompt_messages=[ping_message],
-                           model_parameters={},
+                           model_parameters=required_params,
                           stream=False)
        
        except ClientError as ex:
@@ -297,7 +297,6 @@ class CohereLargeLanguageModel(LargeLanguageModel):
            chat_history=chat_histories,
            model=real_model,
            stream=stream,
-            return_preamble=True,
            **model_parameters,
        )

@@ -0,0 +1,4 @@
+model: jina-colbert-v1-en
+model_type: rerank
+model_properties:
+  context_size: 8192
@@ -1,4 +1,3 @@
-import importlib
 import logging
 import os
 from typing import Optional
@@ -10,6 +9,7 @@ from core.model_runtime.entities.provider_entities import ProviderConfig, Provid
 from core.model_runtime.model_providers.__base.model_provider import ModelProvider
 from core.model_runtime.schema_validators.model_credential_schema_validator import ModelCredentialSchemaValidator
 from core.model_runtime.schema_validators.provider_credential_schema_validator import ProviderCredentialSchemaValidator
+from core.utils.module_import_helper import load_single_subclass_from_source
 from core.utils.position_helper import get_position_map, sort_to_dict_by_position_map

 logger = logging.getLogger(__name__)
@@ -229,15 +229,10 @@ class ModelProviderFactory:

            # Dynamic loading {model_provider_name}.py file and find the subclass of ModelProvider
            py_path = os.path.join(model_provider_dir_path, model_provider_name + '.py')
-            spec = importlib.util.spec_from_file_location(f'core.model_runtime.model_providers.{model_provider_name}.{model_provider_name}', py_path)
-            mod = importlib.util.module_from_spec(spec)
-            spec.loader.exec_module(mod)
-
-            model_provider_class = None
-            for name, obj in vars(mod).items():
-                if isinstance(obj, type) and issubclass(obj, ModelProvider) and obj != ModelProvider:
-                    model_provider_class = obj
-                    break
+            model_provider_class = load_single_subclass_from_source(
+                module_name=f'core.model_runtime.model_providers.{model_provider_name}.{model_provider_name}',
+                script_path=py_path,
+                parent_type=ModelProvider)

            if not model_provider_class:
                logger.warning(f"Missing Model Provider Class that extends ModelProvider in {py_path}, Skip.")
@@ -1,8 +1,31 @@
+import json
 from collections.abc import Generator
-from typing import Optional, Union
+from typing import Optional, Union, cast

-from core.model_runtime.entities.llm_entities import LLMResult
-from core.model_runtime.entities.message_entities import PromptMessage, PromptMessageTool
+import requests
+
+from core.model_runtime.entities.common_entities import I18nObject
+from core.model_runtime.entities.llm_entities import LLMMode, LLMResult, LLMResultChunk, LLMResultChunkDelta
+from core.model_runtime.entities.message_entities import (
+    AssistantPromptMessage,
+    ImagePromptMessageContent,
+    PromptMessage,
+    PromptMessageContent,
+    PromptMessageContentType,
+    PromptMessageTool,
+    SystemPromptMessage,
+    ToolPromptMessage,
+    UserPromptMessage,
+)
+from core.model_runtime.entities.model_entities import (
+    AIModelEntity,
+    FetchFrom,
+    ModelFeature,
+    ModelPropertyKey,
+    ModelType,
+    ParameterRule,
+    ParameterType,
+)
 from core.model_runtime.model_providers.openai_api_compatible.llm.llm import OAIAPICompatLargeLanguageModel


@@ -13,6 +36,7 @@ class MoonshotLargeLanguageModel(OAIAPICompatLargeLanguageModel):
                stream: bool = True, user: Optional[str] = None) \
            -> Union[LLMResult, Generator]:
        self._add_custom_parameters(credentials)
+        self._add_function_call(model, credentials)
        user = user[:32] if user else None
        return super()._invoke(model, credentials, prompt_messages, model_parameters, tools, stop, stream, user)

@@ -20,7 +44,293 @@ class MoonshotLargeLanguageModel(OAIAPICompatLargeLanguageModel):
        self._add_custom_parameters(credentials)
        super().validate_credentials(model, credentials)

-    @staticmethod
-    def _add_custom_parameters(credentials: dict) -> None:
+    def get_customizable_model_schema(self, model: str, credentials: dict) -> AIModelEntity | None:
+        return AIModelEntity(
+            model=model,
+            label=I18nObject(en_US=model, zh_Hans=model),
+            model_type=ModelType.LLM,
+            features=[ModelFeature.TOOL_CALL, ModelFeature.MULTI_TOOL_CALL, ModelFeature.STREAM_TOOL_CALL] 
+                if credentials.get('function_calling_type') == 'tool_call' 
+                else [],
+            fetch_from=FetchFrom.CUSTOMIZABLE_MODEL,
+            model_properties={
+                ModelPropertyKey.CONTEXT_SIZE: int(credentials.get('context_size', 4096)),
+                ModelPropertyKey.MODE: LLMMode.CHAT.value,
+            },
+            parameter_rules=[
+                ParameterRule(
+                    name='temperature',
+                    use_template='temperature',
+                    label=I18nObject(en_US='Temperature', zh_Hans='温度'),
+                    type=ParameterType.FLOAT,
+                ),
+                ParameterRule(
+                    name='max_tokens',
+                    use_template='max_tokens',
+                    default=512,
+                    min=1,
+                    max=int(credentials.get('max_tokens', 4096)),
+                    label=I18nObject(en_US='Max Tokens', zh_Hans='最大标记'),
+                    type=ParameterType.INT,
+                ),
+                ParameterRule(
+                    name='top_p',
+                    use_template='top_p',
+                    label=I18nObject(en_US='Top P', zh_Hans='Top P'),
+                    type=ParameterType.FLOAT,
+                ),
+            ]
+        )
+
+    def _add_custom_parameters(self, credentials: dict) -> None:
        credentials['mode'] = 'chat'
        credentials['endpoint_url'] = 'https://api.moonshot.cn/v1'
+
+    def _add_function_call(self, model: str, credentials: dict) -> None:
+        model_schema = self.get_model_schema(model, credentials)
+        if model_schema and set([
+            ModelFeature.TOOL_CALL, ModelFeature.MULTI_TOOL_CALL
+        ]).intersection(model_schema.features or []):
+            credentials['function_calling_type'] = 'tool_call'
+
+    def _convert_prompt_message_to_dict(self, message: PromptMessage) -> dict:
+        """
+        Convert PromptMessage to dict for OpenAI API format
+        """
+        if isinstance(message, UserPromptMessage):
+            message = cast(UserPromptMessage, message)
+            if isinstance(message.content, str):
+                message_dict = {"role": "user", "content": message.content}
+            else:
+                sub_messages = []
+                for message_content in message.content:
+                    if message_content.type == PromptMessageContentType.TEXT:
+                        message_content = cast(PromptMessageContent, message_content)
+                        sub_message_dict = {
+                            "type": "text",
+                            "text": message_content.data
+                        }
+                        sub_messages.append(sub_message_dict)
+                    elif message_content.type == PromptMessageContentType.IMAGE:
+                        message_content = cast(ImagePromptMessageContent, message_content)
+                        sub_message_dict = {
+                            "type": "image_url",
+                            "image_url": {
+                                "url": message_content.data,
+                                "detail": message_content.detail.value
+                            }
+                        }
+                        sub_messages.append(sub_message_dict)
+                message_dict = {"role": "user", "content": sub_messages}
+        elif isinstance(message, AssistantPromptMessage):
+            message = cast(AssistantPromptMessage, message)
+            message_dict = {"role": "assistant", "content": message.content}
+            if message.tool_calls:
+                message_dict["tool_calls"] = []
+                for function_call in message.tool_calls:
+                    message_dict["tool_calls"].append({
+                        "id": function_call.id,
+                        "type": function_call.type,
+                        "function": {
+                            "name": f"functions.{function_call.function.name}",
+                            "arguments": function_call.function.arguments
+                        }
+                    })
+        elif isinstance(message, ToolPromptMessage):
+            message = cast(ToolPromptMessage, message)
+            message_dict = {"role": "tool", "content": message.content, "tool_call_id": message.tool_call_id}
+            if not message.name.startswith("functions."):
+                message.name = f"functions.{message.name}"
+        elif isinstance(message, SystemPromptMessage):
+            message = cast(SystemPromptMessage, message)
+            message_dict = {"role": "system", "content": message.content}
+        else:
+            raise ValueError(f"Got unknown type {message}")
+
+        if message.name:
+            message_dict["name"] = message.name
+
+        return message_dict
+
+    def _extract_response_tool_calls(self, response_tool_calls: list[dict]) -> list[AssistantPromptMessage.ToolCall]:
+        """
+        Extract tool calls from response
+
+        :param response_tool_calls: response tool calls
+        :return: list of tool calls
+        """
+        tool_calls = []
+        if response_tool_calls:
+            for response_tool_call in response_tool_calls:
+                function = AssistantPromptMessage.ToolCall.ToolCallFunction(
+                    name=response_tool_call["function"]["name"] if response_tool_call.get("function", {}).get("name") else "",
+                    arguments=response_tool_call["function"]["arguments"] if response_tool_call.get("function", {}).get("arguments") else ""
+                )
+
+                tool_call = AssistantPromptMessage.ToolCall(
+                    id=response_tool_call["id"] if response_tool_call.get("id") else "",
+                    type=response_tool_call["type"] if response_tool_call.get("type") else "",
+                    function=function
+                )
+                tool_calls.append(tool_call)
+
+        return tool_calls
+
+    def _handle_generate_stream_response(self, model: str, credentials: dict, response: requests.Response,
+                                         prompt_messages: list[PromptMessage]) -> Generator:
+        """
+        Handle llm stream response
+
+        :param model: model name
+        :param credentials: model credentials
+        :param response: streamed response
+        :param prompt_messages: prompt messages
+        :return: llm response chunk generator
+        """
+        full_assistant_content = ''
+        chunk_index = 0
+
+        def create_final_llm_result_chunk(index: int, message: AssistantPromptMessage, finish_reason: str) \
+                -> LLMResultChunk:
+            # calculate num tokens
+            prompt_tokens = self._num_tokens_from_string(model, prompt_messages[0].content)
+            completion_tokens = self._num_tokens_from_string(model, full_assistant_content)
+
+            # transform usage
+            usage = self._calc_response_usage(model, credentials, prompt_tokens, completion_tokens)
+
+            return LLMResultChunk(
+                model=model,
+                prompt_messages=prompt_messages,
+                delta=LLMResultChunkDelta(
+                    index=index,
+                    message=message,
+                    finish_reason=finish_reason,
+                    usage=usage
+                )
+            )
+
+        tools_calls: list[AssistantPromptMessage.ToolCall] = []
+        finish_reason = "Unknown"
+
+        def increase_tool_call(new_tool_calls: list[AssistantPromptMessage.ToolCall]):
+            def get_tool_call(tool_name: str):
+                if not tool_name:
+                    return tools_calls[-1]
+
+                tool_call = next((tool_call for tool_call in tools_calls if tool_call.function.name == tool_name), None)
+                if tool_call is None:
+                    tool_call = AssistantPromptMessage.ToolCall(
+                        id='',
+                        type='',
+                        function=AssistantPromptMessage.ToolCall.ToolCallFunction(name=tool_name, arguments="")
+                    )
+                    tools_calls.append(tool_call)
+
+                return tool_call
+
+            for new_tool_call in new_tool_calls:
+                # get tool call
+                tool_call = get_tool_call(new_tool_call.function.name)
+                # update tool call
+                if new_tool_call.id:
+                    tool_call.id = new_tool_call.id
+                if new_tool_call.type:
+                    tool_call.type = new_tool_call.type
+                if new_tool_call.function.name:
+                    # remove the functions. prefix
+                    if new_tool_call.function.name.startswith('functions.'):
+                        parts = new_tool_call.function.name.split('functions.')
+                        if len(parts) > 1:
+                            new_tool_call.function.name = parts[1]
+                    tool_call.function.name = new_tool_call.function.name
+                if new_tool_call.function.arguments:
+                    tool_call.function.arguments += new_tool_call.function.arguments
+
+        for chunk in response.iter_lines(decode_unicode=True, delimiter="\n\n"):
+            if chunk:
+                # ignore sse comments
+                if chunk.startswith(':'):
+                    continue
+                decoded_chunk = chunk.strip().lstrip('data: ').lstrip()
+                chunk_json = None
+                try:
+                    chunk_json = json.loads(decoded_chunk)
+                # stream ended
+                except json.JSONDecodeError as e:
+                    yield create_final_llm_result_chunk(
+                        index=chunk_index + 1,
+                        message=AssistantPromptMessage(content=""),
+                        finish_reason="Non-JSON encountered."
+                    )
+                    break
+                if not chunk_json or len(chunk_json['choices']) == 0:
+                    continue
+
+                choice = chunk_json['choices'][0]
+                finish_reason = chunk_json['choices'][0].get('finish_reason')
+                chunk_index += 1
+
+                if 'delta' in choice:
+                    delta = choice['delta']
+                    delta_content = delta.get('content')
+
+                    assistant_message_tool_calls = delta.get('tool_calls', None)
+                    # assistant_message_function_call = delta.delta.function_call
+
+                    # extract tool calls from response
+                    if assistant_message_tool_calls:
+                        tool_calls = self._extract_response_tool_calls(assistant_message_tool_calls)
+                        increase_tool_call(tool_calls)
+
+                    if delta_content is None or delta_content == '':
+                        continue
+
+                    # transform assistant message to prompt message
+                    assistant_prompt_message = AssistantPromptMessage(
+                        content=delta_content,
+                        tool_calls=tool_calls if assistant_message_tool_calls else []
+                    )
+
+                    full_assistant_content += delta_content
+                elif 'text' in choice:
+                    choice_text = choice.get('text', '')
+                    if choice_text == '':
+                        continue
+
+                    # transform assistant message to prompt message
+                    assistant_prompt_message = AssistantPromptMessage(content=choice_text)
+                    full_assistant_content += choice_text
+                else:
+                    continue
+
+                # check payload indicator for completion
+                yield LLMResultChunk(
+                    model=model,
+                    prompt_messages=prompt_messages,
+                    delta=LLMResultChunkDelta(
+                        index=chunk_index,
+                        message=assistant_prompt_message,
+                    )
+                )
+
+            chunk_index += 1
+        
+        if tools_calls:
+            yield LLMResultChunk(
+                model=model,
+                prompt_messages=prompt_messages,
+                delta=LLMResultChunkDelta(
+                    index=chunk_index,
+                    message=AssistantPromptMessage(
+                        tool_calls=tools_calls,
+                        content=""
+                    ),
+                )
+            )
+
+        yield create_final_llm_result_chunk(
+            index=chunk_index,
+            message=AssistantPromptMessage(content=""),
+            finish_reason=finish_reason
+        )
@@ -20,6 +20,7 @@ supported_model_types:
  - llm
 configurate_methods:
  - predefined-model
+  - customizable-model
 provider_credential_schema:
  credential_form_schemas:
    - variable: api_key
@@ -30,3 +31,51 @@ provider_credential_schema:
      placeholder:
        zh_Hans: 在此输入您的 API Key
        en_US: Enter your API Key
+model_credential_schema:
+  model:
+    label:
+      en_US: Model Name
+      zh_Hans: 模型名称
+    placeholder:
+      en_US: Enter your model name
+      zh_Hans: 输入模型名称
+  credential_form_schemas:
+    - variable: api_key
+      label:
+        en_US: API Key
+      type: secret-input
+      required: true
+      placeholder:
+        zh_Hans: 在此输入您的 API Key
+        en_US: Enter your API Key
+    - variable: context_size
+      label:
+        zh_Hans: 模型上下文长度
+        en_US: Model context size
+      required: true
+      type: text-input
+      default: '4096'
+      placeholder:
+        zh_Hans: 在此输入您的模型上下文长度
+        en_US: Enter your Model context size
+    - variable: max_tokens
+      label:
+        zh_Hans: 最大 token 上限
+        en_US: Upper bound for max tokens
+      default: '4096'
+      type: text-input
+    - variable: function_calling_type
+      label:
+        en_US: Function calling
+      type: select
+      required: false
+      default: no_call
+      options:
+        - value: no_call
+          label:
+            en_US: Not supported
+            zh_Hans: 不支持
+        - value: tool_call
+          label:
+            en_US: Tool Call
+            zh_Hans: Tool Call
@@ -378,6 +378,34 @@ class OAIAPICompatLargeLanguageModel(_CommonOAI_API_Compat, LargeLanguageModel):
        delimiter = credentials.get("stream_mode_delimiter", "\n\n")
        delimiter = codecs.decode(delimiter, "unicode_escape")

+        tools_calls: list[AssistantPromptMessage.ToolCall] = []
+
+        def increase_tool_call(new_tool_calls: list[AssistantPromptMessage.ToolCall]):
+            def get_tool_call(tool_call_id: str):
+                tool_call = next(
+                    (tool_call for tool_call in tools_calls if tool_call.id == tool_call_id), None
+                )
+                if tool_call is None:
+                    tool_call = AssistantPromptMessage.ToolCall(
+                        id='', 
+                        type='function', 
+                        function=AssistantPromptMessage.ToolCall.ToolCallFunction(
+                            name='',
+                            arguments=''
+                        )
+                    )
+                    tools_calls.append(tool_call)
+                return tool_call
+
+            for new_tool_call in new_tool_calls:
+                # get tool call
+                tool_call = get_tool_call(new_tool_call.id)
+                # update tool call
+                tool_call.id = new_tool_call.id
+                tool_call.type = new_tool_call.type
+                tool_call.function.name = new_tool_call.function.name
+                tool_call.function.arguments += new_tool_call.function.arguments
+
        for chunk in response.iter_lines(decode_unicode=True, delimiter=delimiter):
            if chunk:
                # ignore sse comments
@@ -405,8 +433,6 @@ class OAIAPICompatLargeLanguageModel(_CommonOAI_API_Compat, LargeLanguageModel):
                if 'delta' in choice:
                    delta = choice['delta']
                    delta_content = delta.get('content')
-                    if delta_content is None or delta_content == '':
-                        continue

                    assistant_message_tool_calls = delta.get('tool_calls', None)
                    # assistant_message_function_call = delta.delta.function_call
@@ -414,6 +440,11 @@ class OAIAPICompatLargeLanguageModel(_CommonOAI_API_Compat, LargeLanguageModel):
                    # extract tool calls from response
                    if assistant_message_tool_calls:
                        tool_calls = self._extract_response_tool_calls(assistant_message_tool_calls)
+                        increase_tool_call(tool_calls)
+
+                    if delta_content is None or delta_content == '':
+                        continue
+
                    # function_call = self._extract_response_function_call(assistant_message_function_call)
                    # tool_calls = [function_call] if function_call else []

@@ -437,6 +468,18 @@ class OAIAPICompatLargeLanguageModel(_CommonOAI_API_Compat, LargeLanguageModel):

                # check payload indicator for completion
                if finish_reason is not None:
+                    yield LLMResultChunk(
+                        model=model,
+                        prompt_messages=prompt_messages,
+                        delta=LLMResultChunkDelta(
+                            index=chunk_index,
+                            message=AssistantPromptMessage(
+                                tool_calls=tools_calls,
+                            ),
+                            finish_reason=finish_reason
+                        )
+                    )
+
                    yield create_final_llm_result_chunk(
                        index=chunk_index,
                        message=assistant_prompt_message,
@@ -735,4 +778,4 @@ class OAIAPICompatLargeLanguageModel(_CommonOAI_API_Compat, LargeLanguageModel):
                function=function
            )

-        return tool_call
+        return tool_call
@@ -0,0 +1,10 @@
+<svg width="25" height="21" viewBox="0 0 25 21" fill="none" xmlns="http://www.w3.org/2000/svg">
+<path d="M1.05858 10.1738C1.76158 10.1738 4.47988 9.56715 5.88589 8.77041C7.2919 7.97367 7.2919 7.97367 10.1977 5.91152C13.8766 3.30069 16.4779 4.17486 20.7428 4.17486" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M11.4182 7.63145L11.3787 7.65951C8.50565 9.69845 8.42504 9.75566 6.92566 10.6053C5.98567 11.138 4.74704 11.5436 3.75151 11.8089C2.80313 12.0615 1.71203 12.2829 1.05858 12.2829V8.06483C1.05075 8.06483 1.05422 8.06445 1.06984 8.06276C1.11491 8.05788 1.26116 8.04203 1.52896 7.9926C1.84599 7.9341 2.24205 7.84582 2.6657 7.73296C3.55657 7.49564 4.3801 7.1996 4.84612 6.93552C4.88175 6.91533 4.91635 6.89573 4.95001 6.87666C6.15007 6.19693 6.15657 6.19325 8.97708 4.1916C12.5199 1.67735 15.5815 1.83587 18.5849 1.99138C19.3056 2.0287 20.0229 2.06584 20.7428 2.06584V6.28388C19.6102 6.28388 18.6583 6.24193 17.8263 6.20527C15.1245 6.08621 13.685 6.02278 11.4182 7.63145Z" fill="black"/>
+<path d="M24.8671 4.20087L17.6613 8.36117V0.0405881L24.8671 4.20087Z" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M17.6378 0L24.9139 4.20087L17.6378 8.40176V0ZM17.6847 0.0811762V8.32058L24.8202 4.20087L17.6847 0.0811762Z" fill="black"/>
+<path d="M0.917975 10.1764C1.62098 10.1764 4.33927 10.7831 5.74529 11.5799C7.1513 12.3766 7.1513 12.3766 10.0571 14.4388C13.736 17.0496 16.3373 16.1754 20.6022 16.1754" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M0.929234 12.2875C0.913615 12.2858 0.910145 12.2854 0.917975 12.2854V8.06741C1.57142 8.06741 2.66253 8.28878 3.61091 8.54142C4.60644 8.80663 5.84507 9.21231 6.78506 9.74497C8.28444 10.5946 8.36505 10.6518 11.2381 12.6908L11.2776 12.7188C13.5444 14.3275 14.9839 14.2641 17.6857 14.145C18.5177 14.1083 19.4696 14.0664 20.6022 14.0664V18.2844C19.8823 18.2844 19.165 18.3216 18.4443 18.3589C15.4409 18.5144 12.3793 18.6729 8.83648 16.1587C6.01597 14.157 6.00947 14.1533 4.80941 13.4736C4.77575 13.4545 4.74115 13.4349 4.70551 13.4148C4.2395 13.1507 3.41597 12.8546 2.5251 12.6173C2.10145 12.5045 1.70538 12.4162 1.38836 12.3577C1.12056 12.3083 0.974309 12.2924 0.929234 12.2875Z" fill="black"/>
+<path d="M24.7265 16.1494L17.5207 11.9892V20.3097L24.7265 16.1494Z" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M17.4972 11.9486L24.7733 16.1494L17.4972 20.3503V11.9486ZM17.5441 12.0297V20.2691L24.6796 16.1494L17.5441 12.0297Z" fill="black"/>
+</svg>
@@ -0,0 +1,46 @@
+from collections.abc import Generator
+from typing import Optional, Union
+
+from core.model_runtime.entities.llm_entities import LLMResult
+from core.model_runtime.entities.message_entities import PromptMessage, PromptMessageTool
+from core.model_runtime.entities.model_entities import AIModelEntity
+from core.model_runtime.model_providers.openai_api_compatible.llm.llm import OAIAPICompatLargeLanguageModel
+
+
+class OpenRouterLargeLanguageModel(OAIAPICompatLargeLanguageModel):
+
+    def _update_endpoint_url(self, credentials: dict):
+        credentials['endpoint_url'] = "https://openrouter.ai/api/v1"
+        return credentials
+
+    def _invoke(self, model: str, credentials: dict,
+                prompt_messages: list[PromptMessage], model_parameters: dict,
+                tools: Optional[list[PromptMessageTool]] = None, stop: Optional[list[str]] = None,
+                stream: bool = True, user: Optional[str] = None) \
+            -> Union[LLMResult, Generator]:
+        cred_with_endpoint = self._update_endpoint_url(credentials=credentials)
+
+        return super()._invoke(model, cred_with_endpoint, prompt_messages, model_parameters, tools, stop, stream, user)
+
+    def validate_credentials(self, model: str, credentials: dict) -> None:
+        cred_with_endpoint = self._update_endpoint_url(credentials=credentials)
+
+        return super().validate_credentials(model, cred_with_endpoint)
+
+    def _generate(self, model: str, credentials: dict, prompt_messages: list[PromptMessage], model_parameters: dict,
+                  tools: Optional[list[PromptMessageTool]] = None, stop: Optional[list[str]] = None,
+                  stream: bool = True, user: Optional[str] = None) -> Union[LLMResult, Generator]:
+        cred_with_endpoint = self._update_endpoint_url(credentials=credentials)
+
+        return super()._generate(model, cred_with_endpoint, prompt_messages, model_parameters, tools, stop, stream, user)
+
+    def get_customizable_model_schema(self, model: str, credentials: dict) -> AIModelEntity:
+        cred_with_endpoint = self._update_endpoint_url(credentials=credentials)
+
+        return super().get_customizable_model_schema(model, cred_with_endpoint)
+
+    def get_num_tokens(self, model: str, credentials: dict, prompt_messages: list[PromptMessage],
+                       tools: Optional[list[PromptMessageTool]] = None) -> int:
+        cred_with_endpoint = self._update_endpoint_url(credentials=credentials)
+
+        return super().get_num_tokens(model, cred_with_endpoint, prompt_messages, tools)
@@ -0,0 +1,11 @@
+import logging
+
+from core.model_runtime.model_providers.__base.model_provider import ModelProvider
+
+logger = logging.getLogger(__name__)
+
+
+class OpenRouterProvider(ModelProvider):
+
+    def validate_provider_credentials(self, credentials: dict) -> None:
+        pass
@@ -0,0 +1,75 @@
+provider: openrouter
+label:
+  en_US: openrouter.ai
+icon_small:
+  en_US: openrouter_square.svg
+icon_large:
+  en_US: openrouter.svg
+background: "#F1EFED"
+help:
+  title:
+    en_US: Get your API key from openrouter.ai
+    zh_Hans: 从 openrouter.ai 获取 API Key
+  url:
+    en_US: https://openrouter.ai/keys
+supported_model_types:
+  - llm
+configurate_methods:
+  - customizable-model
+model_credential_schema:
+  model:
+    label:
+      en_US: Model Name
+      zh_Hans: 模型名称
+    placeholder:
+      en_US: Enter full model name
+      zh_Hans: 输入模型全称
+  credential_form_schemas:
+    - variable: api_key
+      required: true
+      label:
+        en_US: API Key
+      type: secret-input
+      placeholder:
+        zh_Hans: 在此输入您的 API Key
+        en_US: Enter your API Key
+    - variable: mode
+      show_on:
+        - variable: __model_type
+          value: llm
+      label:
+        en_US: Completion mode
+      type: select
+      required: false
+      default: chat
+      placeholder:
+        zh_Hans: 选择对话类型
+        en_US: Select completion mode
+      options:
+        - value: completion
+          label:
+            en_US: Completion
+            zh_Hans: 补全
+        - value: chat
+          label:
+            en_US: Chat
+            zh_Hans: 对话
+    - variable: context_size
+      label:
+        zh_Hans: 模型上下文长度
+        en_US: Model context size
+      required: true
+      type: text-input
+      default: "4096"
+      placeholder:
+        zh_Hans: 在此输入您的模型上下文长度
+        en_US: Enter your Model context size
+    - variable: max_tokens_to_sample
+      label:
+        zh_Hans: 最大 token 上限
+        en_US: Upper bound for max tokens
+      show_on:
+        - variable: __model_type
+          value: llm
+      default: "4096"
+      type: text-input
@@ -148,7 +148,8 @@ class SparkLLMClient:
        data = {
            "header": {
                "app_id": self.app_id,
-                "uid": user_id
+                # resolve this error message => $.header.uid' length must be less or equal than 32
+                "uid": user_id[:32] if user_id else None
            },
            "parameter": {
                "chat": {
@@ -0,0 +1,37 @@
+model: ernie-3.5-4k-0205
+label:
+  en_US: Ernie-3.5-4k-0205
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 4096
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0.1
+    max: 1.0
+    default: 0.8
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 2
+    max: 2048
+  - name: presence_penalty
+    use_template: presence_penalty
+  - name: frequency_penalty
+    use_template: frequency_penalty
+  - name: response_format
+    use_template: response_format
+  - name: disable_search
+    label:
+      zh_Hans: 禁用搜索
+      en_US: Disable Search
+    type: boolean
+    help:
+      zh_Hans: 禁用模型自行进行外部搜索。
+      en_US: Disable the model to perform external search.
+    required: false
@@ -0,0 +1,37 @@
+model: ernie-3.5-8k-0205
+label:
+  en_US: Ernie-3.5-8K-0205
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 8192
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0.1
+    max: 1.0
+    default: 0.8
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 2
+    max: 2048
+  - name: presence_penalty
+    use_template: presence_penalty
+  - name: frequency_penalty
+    use_template: frequency_penalty
+  - name: response_format
+    use_template: response_format
+  - name: disable_search
+    label:
+      zh_Hans: 禁用搜索
+      en_US: Disable Search
+    type: boolean
+    help:
+      zh_Hans: 禁用模型自行进行外部搜索。
+      en_US: Disable the model to perform external search.
+    required: false
@@ -0,0 +1,37 @@
+model: ernie-3.5-8k-1222
+label:
+  en_US: Ernie-3.5-8K-1222
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 8192
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0.1
+    max: 1.0
+    default: 0.8
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 2
+    max: 2048
+  - name: presence_penalty
+    use_template: presence_penalty
+  - name: frequency_penalty
+    use_template: frequency_penalty
+  - name: response_format
+    use_template: response_format
+  - name: disable_search
+    label:
+      zh_Hans: 禁用搜索
+      en_US: Disable Search
+    type: boolean
+    help:
+      zh_Hans: 禁用模型自行进行外部搜索。
+      en_US: Disable the model to perform external search.
+    required: false
@@ -0,0 +1,37 @@
+model: ernie-3.5-8k
+label:
+  en_US: Ernie-3.5-8K
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 8192
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0.1
+    max: 1.0
+    default: 0.8
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 2
+    max: 2048
+  - name: presence_penalty
+    use_template: presence_penalty
+  - name: frequency_penalty
+    use_template: frequency_penalty
+  - name: response_format
+    use_template: response_format
+  - name: disable_search
+    label:
+      zh_Hans: 禁用搜索
+      en_US: Disable Search
+    type: boolean
+    help:
+      zh_Hans: 禁用模型自行进行外部搜索。
+      en_US: Disable the model to perform external search.
+    required: false
@@ -0,0 +1,37 @@
+model: ernie-4.0-8k
+label:
+  en_US: Ernie-4.0-8K
+model_type: llm
+features:
+  - agent-thought
+model_properties:
+  mode: chat
+  context_size: 8192
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+    min: 0.1
+    max: 1.0
+    default: 0.8
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 2
+    max: 2048
+  - name: presence_penalty
+    use_template: presence_penalty
+  - name: frequency_penalty
+    use_template: frequency_penalty
+  - name: response_format
+    use_template: response_format
+  - name: disable_search
+    label:
+      zh_Hans: 禁用搜索
+      en_US: Disable Search
+    type: boolean
+    help:
+      zh_Hans: 禁用模型自行进行外部搜索。
+      en_US: Disable the model to perform external search.
+    required: false
@@ -36,3 +36,4 @@ parameter_rules:
      zh_Hans: 禁用模型自行进行外部搜索。
      en_US: Disable the model to perform external search.
    required: false
+deprecated: true
@@ -36,3 +36,4 @@ parameter_rules:
      zh_Hans: 禁用模型自行进行外部搜索。
      en_US: Disable the model to perform external search.
    required: false
+deprecated: true
@@ -27,3 +27,4 @@ parameter_rules:
    use_template: frequency_penalty
  - name: response_format
    use_template: response_format
+deprecated: true
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
takatostandGitHub	74de7cf33c	version to 0.6.1 (#3253 )	2024-04-09 21:21:09 +08:00
crazywoolaandGitHub	f5e65b98a9	feat: remove unregistered-llm-in-debug (#3251 )	2024-04-09 20:49:52 +08:00
Chenhe GuandGitHub	eb76d7a226	make sure validation flow works for all model providers in bedrock (#3250 )	2024-04-09 20:42:18 +08:00
YeuolyandGitHub	e635f3dc1d	chore: remove langchain in tools (#3247 )	2024-04-09 19:28:22 +08:00
takatostandGitHub	2a6b7d57cb	fix: token is not logging of question classifier node (#3249 )	2024-04-09 19:25:08 +08:00
JoelandGitHub	8bb225bec6	fix: number type in app would render as select type in webapp (#3244 )	2024-04-09 18:33:47 +08:00
zxhlyhandGitHub	39d3fc4742	feat: prompt-editor support undo (#3242 )	2024-04-09 18:18:53 +08:00
KVOJJJinandGitHub	5c98260cec	Fix: picture of workflow (#3241 )	2024-04-09 17:49:45 +08:00
takatostandGitHub	f599f41336	fix: empty conversation list of explore chatbot (#3235 )	2024-04-09 17:04:48 +08:00
zxhlyhandGitHub	1384a6d0fd	fix: workflow run edge status (#3236 )	2024-04-09 16:53:34 +08:00
Bowen LiangandGitHub	28089c98c1	fix: skip Celery warning by setting broker_connection_retry_on_startup config (#3188 )	2024-04-09 16:14:43 +08:00
minakokojimaandGitHub	10d6d50b6c	update link (#3226 )	2024-04-09 16:09:46 +08:00
JoelandGitHub	752f6fb15a	fix: file not uploaded caused api error (#3228 )	2024-04-09 15:54:36 +08:00
JyongandGitHub	8fcf459285	fix milvus database name parameter missed (#3229 )	2024-04-09 15:54:13 +08:00
Leo QandGitHub	9c01bcb3e5	feat: support setting database used in Milvus (#3003 )	2024-04-09 15:39:36 +08:00
YeuolyandGitHub	a2c068d949	feat: moonshot function call (#3227 )	2024-04-09 15:30:09 +08:00
takatostandGitHub	4ad3f2cdc2	fix: image text when retrieve chat histories (#3220 )	2024-04-09 15:20:45 +08:00
legaoandGitHub	29918c498c	fixed the issue of missing cleanup function in the AudioBtn component (#3133 )	2024-04-09 15:10:58 +08:00
JoelandGitHub	269432a5e6	fix: vision config doesn't enabled in llm (#3225 )	2024-04-09 15:07:43 +08:00
takatostandGitHub	a33b774314	fix: latest image tag not push in GitHub action (#3213 )	2024-04-09 14:35:39 +08:00
YeuolyandGitHub	cc5ccaaca1	fix: incomplete response (#3215 )	2024-04-09 14:35:25 +08:00
JyongandGitHub	33ea689861	fix detached instance error in keyword index create thread and fix question classifier node out of index error (#3219 )	2024-04-09 14:34:51 +08:00
Chenhe GuandGitHub	581836b716	Update README.md (#3212 )	2024-04-09 14:34:42 +08:00
Bowen LiangandGitHub	0516b78d6f	fix: index number in api/README (#3214 )	2024-04-09 13:59:26 +08:00
JyongandGitHub	84d7cbf916	fix economy index search in workflow (#3205 )	2024-04-09 13:20:51 +08:00
Chenhe GuandGitHub	f514fd2182	Update README.md (#3206 )	2024-04-09 12:30:44 +08:00
zxhlyhandGitHub	86707928d4	fix: node connect self (#3194 )	2024-04-09 12:24:41 +08:00
Eric WangandGitHub	3c3fb3cd3f	fix(code_executor): surrogates not allowed error in jinja2 template (#3191 )	2024-04-09 12:21:03 +08:00
YeuolyandGitHub	337899a03d	Fix/code transform result (#3203 )	2024-04-09 12:20:34 +08:00
JatandGitHub	bae0c071cd	Fix: remove unavailable return_preamble parameter in cohere (#3201 ) Signed-off-by: Jat <[email protected]>	2024-04-09 12:11:53 +08:00
JoelandGitHub	19cb3c7871	fix: sometimes chosed old selected knowledge may overwirte the new knowledge (#3199 )	2024-04-09 11:46:59 +08:00
JyongandGitHub	2e4dec365d	Compatible with unique index conflicts (#3183 )	2024-04-09 02:16:19 +08:00
Chenhe GuandGitHub	ca3e2e6cc0	Update README.md to include workflows (#3180 )	2024-04-09 01:49:19 +08:00
JyongandGitHub	283979fc46	fix keyword index error when storage source is S3 (#3182 )	2024-04-09 01:42:58 +08:00
takatostandGitHub	a81c1ab6ae	version to 0.6.0-fix1 (#3179 )	2024-04-09 00:10:20 +08:00
KVOJJJinandGitHub	48d4d55ecc	Fix: features of agent-chat (#3178 )	2024-04-08 23:53:59 +08:00
zxhlyhandGitHub	b7691f5658	fix: prompt editor variable picker (#3177 )	2024-04-08 23:53:09 +08:00
crazywoolaandGitHub	1382f10433	feat: translations (#3176 )	2024-04-08 23:17:16 +08:00
KVOJJJinandGitHub	d8db728c33	Fix: prompt of expert mode (#3168 )	2024-04-08 21:36:27 +08:00
takatostandGitHub	d2259f20cb	fix: app export dsl not include desc (#3167 )	2024-04-08 21:30:18 +08:00
takatostandGitHub	9720d6b7a5	fix: metadata in generate npe issue (#3166 )	2024-04-08 21:30:03 +08:00
takatost GitHub Joel Yeuoly JzoNg StyleZhang jyong nite-knite jyong	7753ba2d37	FEAT: NEW WORKFLOW ENGINE (#3160 ) Co-authored-by: Joel <[email protected]> Co-authored-by: Yeuoly <[email protected]> Co-authored-by: JzoNg <[email protected]> Co-authored-by: StyleZhang <[email protected]> Co-authored-by: jyong <[email protected]> Co-authored-by: nite-knite <[email protected]> Co-authored-by: jyong <[email protected]>	2024-04-08 18:51:46 +08:00
zxhlyhandGitHub	2fb9850af5	fix: knowledge create display error (#3157 )	2024-04-08 16:40:52 +08:00
Jyong GitHub jyong	9eba6ffdd4	Optimize csv and excel extract (#3155 ) Co-authored-by: jyong <[email protected]>	2024-04-08 16:34:43 +08:00
KVOJJJinandGitHub	762657eeef	Fix: stop indexing status check when api of status checking failed (#3156 )	2024-04-08 16:14:31 +08:00
Richards TuandGitHub	16e3b0484d	Update descriptions in StackExchange Tool (#3043 )	2024-04-08 15:40:41 +08:00
JoelandGitHub	974828222e	fix: chat app sometimes may crash (#3151 )	2024-04-08 14:37:39 +08:00
crazywoola GitHub Chenhe Gu	a9700e61db	Feat/update issue template (#3147 ) Co-authored-by: Chenhe Gu <[email protected]>	2024-04-08 02:46:28 +08:00
Even M GitHub mazhanwen	5a23d570b5	fix: Turn off SWR automatic revalidation when window is focused (#3129 ) Co-authored-by: mazhanwen <[email protected]>	2024-04-07 22:43:44 +08:00
Jyong GitHub jyong	28b1c48235	improve qa generate prompt (#3132 ) Co-authored-by: jyong <[email protected]>	2024-04-07 15:21:11 +08:00
Jyong GitHub jyong	ab9fcbdfb9	Duplicate embedding cache check (#3134 ) Co-authored-by: jyong <[email protected]>	2024-04-07 15:20:58 +08:00
KVOJJJinandGitHub	ef80b3711e	chore: update link of feedback (#3130 )	2024-04-07 13:37:32 +08:00
crazywoolaandGitHub	6672a03e7f	feat: update link (#3121 )	2024-04-06 14:57:07 +08:00
Nanguan LinandGitHub	e7833a070e	chore: replace outdated config in vscode debug settings (#3106 )	2024-04-05 17:49:09 +08:00
YeuolyandGitHub	25b9ac3df4	feat: claude3 tool call (#3111 )	2024-04-05 16:35:59 +09:00
Nanguan LinandGitHub	718ac3f83b	Improve ModelTypeEnum type (#3051 )	2024-04-04 15:54:59 +08:00
Vikey ChenandGitHub	e4f686deb7	fix unstructured api,remove unused parameters (#3056 )	2024-04-03 21:00:20 +08:00
JatandGitHub	d241d66a69	fix typo in readme (#3096 ) Signed-off-by: Jat <[email protected]>	2024-04-03 20:29:02 +08:00
呆萌闷油瓶andGitHub	f92a1be0b6	fix typo (#3098 )	2024-04-03 20:26:21 +08:00
Bowen LiangandGitHub	7cc0d47322	fix: update show names for supported file types of xlsx and docx (#3091 )	2024-04-03 20:26:12 +08:00
Chenhe GuandGitHub	da998d09d7	new readme slogan (#3094 )	2024-04-03 13:39:41 +08:00
Jyong GitHub jyong	5e66a60f1c	add embedding cache and clean embedding cache job (#3087 ) Co-authored-by: jyong <[email protected]>	2024-04-02 20:46:24 +08:00
crazywoolaandGitHub	7f55ea0c53	Chore/move chrome ext (#3085 )	2024-04-02 19:51:02 +08:00
QIN2DIMandGitHub	f7d1d9b8b1	fix(duckduckgo-search): invoke error (#3077 )	2024-04-02 18:40:09 +08:00
Salem KorayemandGitHub	6b4c8e76e6	feat (new llm): add support for openrouter (#3042 )	2024-04-02 18:38:46 +08:00
Jyong GitHub jyong	e12a0c154c	add segment function billing check for SAAS env (#3082 ) Co-authored-by: jyong <[email protected]>	2024-04-02 17:55:49 +08:00
Chenhe GuandGitHub	9c7e99e829	Update README.md (#3081 )	2024-04-02 17:19:21 +08:00
takatostandGitHub	d14ea2ecaa	version to 0.5.11-fix1 (#3073 )	2024-04-02 12:51:29 +08:00
Jyong GitHub jyong	a94d86da6d	add keyword table s3 storage support (#3065 ) Co-authored-by: jyong <[email protected]>	2024-04-01 20:19:30 +08:00
arkii GitHub crazywoola	5e591fc1b7	feat: add Feishu(飞书) tool for sending message to chat group bot via webhook (#3059 ) Co-authored-by: crazywoola <[email protected]>	2024-04-01 18:03:45 +08:00
takatostandGitHub	32e83e00e4	feat: use en-US as fallback recommend app if using unmaintained language (#3063 )	2024-04-01 16:15:59 +08:00
Yash ParmarandGitHub	132269618d	FEAT: Add Brave Search and Trello(12 Tools) Included (#3040 )	2024-04-01 14:53:56 +08:00
Jyong GitHub jyong	84d118de07	add redis lock on create collection in multiple thread mode (#3054 ) Co-authored-by: jyong <[email protected]>	2024-04-01 02:10:41 +08:00
Jyong GitHub jyong	1716ac562c	add clean_unused_datasets_task (#3057 ) Co-authored-by: jyong <[email protected]>	2024-04-01 01:34:21 +08:00
呆萌闷油瓶andGitHub	e215aae39a	feat:xinference audio model support (#3045 )	2024-03-31 12:44:11 +08:00
Nanguan LinandGitHub	12782cad4d	Fix typo (#3041 )	2024-03-31 12:41:16 +08:00
Leo QandGitHub	fc5ed17fe9	provide a bit more info in logs when parsing api schema error (#3026 )	2024-03-30 14:44:50 +08:00
YeuolyandGitHub	94d04934b3	fix: agent tool label (#3039 )	2024-03-29 22:15:16 +08:00
takatostandGitHub	1387f9b23e	version to 0.5.11 (#3038 )	2024-03-29 21:09:21 +08:00
takatostandGitHub	6817eab5f1	fix: api / moderation extension import error (#3037 )	2024-03-29 21:07:34 +08:00
zxhlyhandGitHub	218f591a5d	fix: prompt editor linebreak (#3036 )	2024-03-29 21:01:04 +08:00
Richards Tu GitHub crazywoola	17af0de7b6	Add New Tool: StackExchange (#3034 ) Co-authored-by: crazywoola <[email protected]>	2024-03-29 20:28:21 +08:00
Chenhe Gu GitHub crazywoola crazywoola	9d962053a2	Fix claude request errors in bedrock (#3015 ) Co-authored-by: crazywoola <[email protected]> Co-authored-by: crazywoola <[email protected]>	2024-03-29 13:57:45 +08:00
kun321andGitHub	59909b5ca7	update the discord Invalid invite (#3028 )	2024-03-29 13:16:52 +08:00
Jyong GitHub jyong	a6cd0f0e73	fix add segment when dataset and document is empty (#3021 ) Co-authored-by: jyong <[email protected]>	2024-03-29 13:06:00 +08:00
Richards TuandGitHub	2c43393bf1	Add New Tool: DevDocs (#2993 )	2024-03-29 11:21:02 +08:00
Jyong GitHub jyong	669c8c3cca	some optimization for admin api key, create tenant and reset-encrypt-key-pair command (#3013 ) Co-authored-by: jyong <[email protected]>	2024-03-28 17:02:52 +08:00
Jyong GitHub jyong	b0b0cc045f	add mutil-thread document embedding (#3016 ) Co-authored-by: jyong <[email protected]>	2024-03-28 17:02:35 +08:00
crazywoolaandGitHub	20d16d7b31	doc: update helm charts (#3012 )	2024-03-28 13:02:41 +08:00
RickyandGitHub	714722bb2d	fix: 'next' button unresponsive when uploading additional documents before previous batch completes (#2991 )	2024-03-28 12:28:15 +08:00
Bowen Liang GitHub takatost	830495a607	bump celery from 5.2 to 5.3 (#2478 ) Co-authored-by: takatost <[email protected]>	2024-03-28 11:53:48 +08:00
Bowen LiangandGitHub	41a4593b6d	bump redis client to 5.0 and enable hiredis support (#2518 )	2024-03-28 11:40:21 +08:00
Bowen LiangandGitHub	08b727833e	generalize helper for loading module from source (#2862 )	2024-03-28 11:37:26 +08:00
aqachunandGitHub	c8b82b9d08	fix: missing comma in JSON for `/completion-messages` request (#2999 )	2024-03-27 14:31:06 +08:00
WeaxsandGitHub	5becb4c43a	update wenxin llm (#2929 )	2024-03-27 11:36:21 +08:00
KennyandGitHub	13694293e3	fix: resolve header.uid' length must be less or equal than 32 on Spark V1.5 (#2983 )	2024-03-27 09:58:41 +08:00
RickyandGitHub	815beac356	Fix the time in the annotation from 12-hour clock to 24-hour clock. (#2990 )	2024-03-27 09:08:38 +08:00
legaoandGitHub	5e60204832	fix: progress bar issue (#2957 )	2024-03-26 17:26:58 +08:00
legaoandGitHub	d2624b13a0	fix: the issue of text overflow in the NavSelector component (#2976 )	2024-03-26 17:22:01 +08:00
zxhlyhandGitHub	61f5de9662	fix: chat scroll (#2981 )	2024-03-26 16:19:41 +08:00
RickyandGitHub	40dbf30784	feat: support new reranker [jina-colbert-v1-en] (#2975 )	2024-03-26 11:34:40 +08:00
RickyandGitHub	afd77c4745	fix: the batch annotaion btn should also be loading when progress status is waiting (#2974 )	2024-03-26 11:05:29 +08:00
listengandGitHub	d70bd4aaa4	fix tool_inputs parse error in message that in CoT(ReAct) agent mode (#2949 )	2024-03-26 11:05:10 +08:00
Yulong WangandGitHub	8e05261588	Fix handling of missing required parameters in ApiTool (#2965 )	2024-03-26 10:53:39 +08:00
Weishan-0andGitHub	a676d4387c	fix: Correct image parameter passing in GLM-4v model API calls (#2948 )	2024-03-26 10:43:20 +08:00
Kenny GitHub crazywoola	08a5afcf9f	feat: update nginx and docker-compose files to support HTTPS. (#2940 ) Co-authored-by: crazywoola <[email protected]>	2024-03-26 10:37:43 +08:00
crazywoolaandGitHub	eeaa3c1643	Fix/2969 add model provider ollama not work (#2973 )	2024-03-26 10:26:34 +08:00
Leo QandGitHub	7c8c233cf4	Add S3_ADDRESS_STYLE configuration option (#2934 )	2024-03-26 10:18:26 +08:00
Bowen LiangandGitHub	129a9850eb	fix: correct response hint for generated image to avoid illusion of regernerated image link (#2962 )	2024-03-26 10:13:35 +08:00
Bowen LiangandGitHub	1f98a4fff3	improve: cache tool icons by setting max-age HTTP header and enable gzip compression SVG icons from backend (#2971 )	2024-03-26 10:11:43 +08:00
RickyandGitHub	58e4702b14	fix: white screen when editing annotaion in log panel (#2968 )	2024-03-26 10:10:14 +08:00
colvin777 GitHub colvin	c60749678b	When disabling the "Annotation Reply" button, the backend reports an error. #2904 (#2933 ) Co-authored-by: colvin <[email protected]>	2024-03-25 22:20:40 +08:00
legaoandGitHub	d5214e4644	reuse layout (#2956 )	2024-03-25 15:13:50 +08:00
legaoandGitHub	52804ca6d1	fix: adjust popup panel's z-index value (#2952 )	2024-03-25 15:09:01 +08:00
orangeclkandGitHub	4fb9606361	fix: max_token default help info improved (#2951 )	2024-03-25 10:07:32 +08:00
orangeclkandGitHub	c534d95972	fix: yi model price correction (#2946 )	2024-03-24 12:10:57 +08:00
Nanguan LinandGitHub	46ccfda493	fix: invalid i18 link in README (#2947 )	2024-03-24 12:10:13 +08:00
orangeclkandGitHub	6dc62334d6	doc: model schema document fix and wording about the model price parameter (#2944 )	2024-03-24 12:06:20 +08:00