Merge branch 'main' into feat/node-execution-retry

2026-05-03 08:58:09 +08:00 · 2024-12-18 15:33:24 +08:00
parent 357769c72e 56434db4f5
commit b411087bb7
30 changed files with 403 additions and 35 deletions
--- a/api/core/model_runtime/model_providers/hunyuan/hunyuan.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/hunyuan.yaml
@ -3,8 +3,8 @@ label:
  zh_Hans: 腾讯混元
  en_US: Hunyuan
 description:
-  en_US: Models provided by Tencent Hunyuan, such as hunyuan-standard, hunyuan-standard-256k, hunyuan-pro and hunyuan-lite.
-  zh_Hans: 腾讯混元提供的模型，例如 hunyuan-standard、 hunyuan-standard-256k, hunyuan-pro 和 hunyuan-lite。
+  en_US: Models provided by Tencent Hunyuan, such as hunyuan-standard, hunyuan-standard-256k, hunyuan-pro, hunyuan-role, hunyuan-large, hunyuan-large-role, hunyuan-turbo-latest, hunyuan-large-longcontext, hunyuan-turbo, hunyuan-vision, hunyuan-turbo-vision, hunyuan-functioncall and hunyuan-lite.
+  zh_Hans: 腾讯混元提供的模型，例如 hunyuan-standard、 hunyuan-standard-256k, hunyuan-pro, hunyuan-role, hunyuan-large, hunyuan-large-role, hunyuan-turbo-latest, hunyuan-large-longcontext, hunyuan-turbo, hunyuan-vision, hunyuan-turbo-vision, hunyuan-functioncall 和 hunyuan-lite。
 icon_small:
  en_US: icon_s_en.png
 icon_large:
--- a/api/core/model_runtime/model_providers/hunyuan/llm/_position.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/llm/_position.yaml
@ -4,3 +4,10 @@
 - hunyuan-pro
 - hunyuan-turbo
 - hunyuan-vision
+- hunyuan-role
+- hunyuan-large
+- hunyuan-large-role
+- hunyuan-large-longcontext
+- hunyuan-turbo-latest
+- hunyuan-turbo-vision
+- hunyuan-functioncall
--- a/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-functioncall.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-functioncall.yaml
@ -0,0 +1,38 @@
+model: hunyuan-functioncall
+label:
+  zh_Hans: hunyuan-functioncall
+  en_US: hunyuan-functioncall
+model_type: llm
+features:
+  - agent-thought
+  - tool-call
+  - multi-tool-call
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 32000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 1
+    max: 32000
+  - name: enable_enhance
+    label:
+      zh_Hans: 功能增强
+      en_US: Enable Enhancement
+    type: boolean
+    help:
+      zh_Hans: 功能增强（如搜索）开关，关闭时将直接由主模型生成回复内容，可以降低响应时延（对于流式输出时的首字时延尤为明显）。但在少数场景里，回复效果可能会下降。
+      en_US: Allow the model to perform external search to enhance the generation results.
+    required: false
+    default: true
+pricing:
+  input: '0.004'
+  output: '0.008'
+  unit: '0.001'
+  currency: RMB
--- a/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-large-longcontext.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-large-longcontext.yaml
@ -0,0 +1,38 @@
+model: hunyuan-large-longcontext
+label:
+  zh_Hans: hunyuan-large-longcontext
+  en_US: hunyuan-large-longcontext
+model_type: llm
+features:
+  - agent-thought
+  - tool-call
+  - multi-tool-call
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 134000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 1
+    max: 134000
+  - name: enable_enhance
+    label:
+      zh_Hans: 功能增强
+      en_US: Enable Enhancement
+    type: boolean
+    help:
+      zh_Hans: 功能增强（如搜索）开关，关闭时将直接由主模型生成回复内容，可以降低响应时延（对于流式输出时的首字时延尤为明显）。但在少数场景里，回复效果可能会下降。
+      en_US: Allow the model to perform external search to enhance the generation results.
+    required: false
+    default: true
+pricing:
+  input: '0.006'
+  output: '0.018'
+  unit: '0.001'
+  currency: RMB
--- a/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-large-role.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-large-role.yaml
@ -0,0 +1,38 @@
+model: hunyuan-large-role
+label:
+  zh_Hans: hunyuan-large-role
+  en_US: hunyuan-large-role
+model_type: llm
+features:
+  - agent-thought
+  - tool-call
+  - multi-tool-call
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 32000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 1
+    max: 32000
+  - name: enable_enhance
+    label:
+      zh_Hans: 功能增强
+      en_US: Enable Enhancement
+    type: boolean
+    help:
+      zh_Hans: 功能增强（如搜索）开关，关闭时将直接由主模型生成回复内容，可以降低响应时延（对于流式输出时的首字时延尤为明显）。但在少数场景里，回复效果可能会下降。
+      en_US: Allow the model to perform external search to enhance the generation results.
+    required: false
+    default: true
+pricing:
+  input: '0.004'
+  output: '0.008'
+  unit: '0.001'
+  currency: RMB
--- a/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-large.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-large.yaml
@ -0,0 +1,38 @@
+model: hunyuan-large
+label:
+  zh_Hans: hunyuan-large
+  en_US: hunyuan-large
+model_type: llm
+features:
+  - agent-thought
+  - tool-call
+  - multi-tool-call
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 32000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 1
+    max: 32000
+  - name: enable_enhance
+    label:
+      zh_Hans: 功能增强
+      en_US: Enable Enhancement
+    type: boolean
+    help:
+      zh_Hans: 功能增强（如搜索）开关，关闭时将直接由主模型生成回复内容，可以降低响应时延（对于流式输出时的首字时延尤为明显）。但在少数场景里，回复效果可能会下降。
+      en_US: Allow the model to perform external search to enhance the generation results.
+    required: false
+    default: true
+pricing:
+  input: '0.004'
+  output: '0.012'
+  unit: '0.001'
+  currency: RMB
--- a/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-role.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-role.yaml
@ -0,0 +1,38 @@
+model: hunyuan-role
+label:
+  zh_Hans: hunyuan-role
+  en_US: hunyuan-role
+model_type: llm
+features:
+  - agent-thought
+  - tool-call
+  - multi-tool-call
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 32000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 1
+    max: 32000
+  - name: enable_enhance
+    label:
+      zh_Hans: 功能增强
+      en_US: Enable Enhancement
+    type: boolean
+    help:
+      zh_Hans: 功能增强（如搜索）开关，关闭时将直接由主模型生成回复内容，可以降低响应时延（对于流式输出时的首字时延尤为明显）。但在少数场景里，回复效果可能会下降。
+      en_US: Allow the model to perform external search to enhance the generation results.
+    required: false
+    default: true
+pricing:
+  input: '0.004'
+  output: '0.008'
+  unit: '0.001'
+  currency: RMB
--- a/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-turbo-latest.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-turbo-latest.yaml
@ -0,0 +1,38 @@
+model: hunyuan-turbo-latest
+label:
+  zh_Hans: hunyuan-turbo-latest
+  en_US: hunyuan-turbo-latest
+model_type: llm
+features:
+  - agent-thought
+  - tool-call
+  - multi-tool-call
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 32000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 1
+    max: 32000
+  - name: enable_enhance
+    label:
+      zh_Hans: 功能增强
+      en_US: Enable Enhancement
+    type: boolean
+    help:
+      zh_Hans: 功能增强（如搜索）开关，关闭时将直接由主模型生成回复内容，可以降低响应时延（对于流式输出时的首字时延尤为明显）。但在少数场景里，回复效果可能会下降。
+      en_US: Allow the model to perform external search to enhance the generation results.
+    required: false
+    default: true
+pricing:
+  input: '0.015'
+  output: '0.05'
+  unit: '0.001'
+  currency: RMB
--- a/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-turbo-vision.yaml
+++ b/api/core/model_runtime/model_providers/hunyuan/llm/hunyuan-turbo-vision.yaml
@ -0,0 +1,39 @@
+model: hunyuan-turbo-vision
+label:
+  zh_Hans: hunyuan-turbo-vision
+  en_US: hunyuan-turbo-vision
+model_type: llm
+features:
+  - agent-thought
+  - tool-call
+  - multi-tool-call
+  - stream-tool-call
+  - vision
+model_properties:
+  mode: chat
+  context_size: 8000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: top_p
+    use_template: top_p
+  - name: max_tokens
+    use_template: max_tokens
+    default: 1024
+    min: 1
+    max: 8000
+  - name: enable_enhance
+    label:
+      zh_Hans: 功能增强
+      en_US: Enable Enhancement
+    type: boolean
+    help:
+      zh_Hans: 功能增强（如搜索）开关，关闭时将直接由主模型生成回复内容，可以降低响应时延（对于流式输出时的首字时延尤为明显）。但在少数场景里，回复效果可能会下降。
+      en_US: Allow the model to perform external search to enhance the generation results.
+    required: false
+    default: true
+pricing:
+  input: '0.08'
+  output: '0.08'
+  unit: '0.001'
+  currency: RMB
--- a/api/core/model_runtime/model_providers/openai/llm/_position.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/_position.yaml
@ -1,4 +1,7 @@
- gpt-4o-audio-preview
+- o1
+- o1-2024-12-17
+- o1-mini
+- o1-mini-2024-09-12
 - gpt-4
 - gpt-4o
 - gpt-4o-2024-05-13
@ -7,10 +10,6 @@
 - chatgpt-4o-latest
 - gpt-4o-mini
 - gpt-4o-mini-2024-07-18
- o1-preview
- o1-preview-2024-09-12
- o1-mini
- o1-mini-2024-09-12
 - gpt-4-turbo
 - gpt-4-turbo-2024-04-09
 - gpt-4-turbo-preview
@ -25,4 +24,7 @@
 - gpt-3.5-turbo-1106
 - gpt-3.5-turbo-0613
 - gpt-3.5-turbo-instruct
+- gpt-4o-audio-preview
+- o1-preview
+- o1-preview-2024-09-12
 - text-davinci-003
--- a/api/core/model_runtime/model_providers/openai/llm/chatgpt-4o-latest.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/chatgpt-4o-latest.yaml
@ -22,7 +22,7 @@ parameter_rules:
    use_template: frequency_penalty
  - name: max_tokens
    use_template: max_tokens
-    default: 512
+    default: 16384
    min: 1
    max: 16384
  - name: response_format
--- a/api/core/model_runtime/model_providers/openai/llm/gpt-4o-2024-05-13.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/gpt-4o-2024-05-13.yaml
@ -22,9 +22,9 @@ parameter_rules:
    use_template: frequency_penalty
  - name: max_tokens
    use_template: max_tokens
-    default: 512
+    default: 16384
    min: 1
-    max: 4096
+    max: 16384
  - name: response_format
    label:
      zh_Hans: 回复格式
--- a/api/core/model_runtime/model_providers/openai/llm/gpt-4o-2024-08-06.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/gpt-4o-2024-08-06.yaml
@ -22,7 +22,7 @@ parameter_rules:
    use_template: frequency_penalty
  - name: max_tokens
    use_template: max_tokens
-    default: 512
+    default: 16384
    min: 1
    max: 16384
  - name: response_format
--- a/api/core/model_runtime/model_providers/openai/llm/gpt-4o-2024-11-20.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/gpt-4o-2024-11-20.yaml
@ -22,7 +22,7 @@ parameter_rules:
    use_template: frequency_penalty
  - name: max_tokens
    use_template: max_tokens
-    default: 512
+    default: 16384
    min: 1
    max: 16384
  - name: response_format
--- a/api/core/model_runtime/model_providers/openai/llm/gpt-4o-audio-preview.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/gpt-4o-audio-preview.yaml
@ -22,9 +22,9 @@ parameter_rules:
    use_template: frequency_penalty
  - name: max_tokens
    use_template: max_tokens
-    default: 512
+    default: 16384
    min: 1
-    max: 4096
+    max: 16384
  - name: response_format
    label:
      zh_Hans: 回复格式
--- a/api/core/model_runtime/model_providers/openai/llm/gpt-4o-mini-2024-07-18.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/gpt-4o-mini-2024-07-18.yaml
@ -22,7 +22,7 @@ parameter_rules:
    use_template: frequency_penalty
  - name: max_tokens
    use_template: max_tokens
-    default: 512
+    default: 16384
    min: 1
    max: 16384
  - name: response_format
--- a/api/core/model_runtime/model_providers/openai/llm/gpt-4o-mini.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/gpt-4o-mini.yaml
@ -22,7 +22,7 @@ parameter_rules:
    use_template: frequency_penalty
  - name: max_tokens
    use_template: max_tokens
-    default: 512
+    default: 16384
    min: 1
    max: 16384
  - name: response_format
--- a/api/core/model_runtime/model_providers/openai/llm/gpt-4o.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/gpt-4o.yaml
@ -22,9 +22,9 @@ parameter_rules:
    use_template: frequency_penalty
  - name: max_tokens
    use_template: max_tokens
-    default: 512
+    default: 16384
    min: 1
-    max: 4096
+    max: 16384
  - name: response_format
    label:
      zh_Hans: 回复格式
@ -38,7 +38,7 @@ parameter_rules:
      - text
      - json_object
 pricing:
-  input: '5.00'
-  output: '15.00'
+  input: '2.50'
+  output: '10.00'
  unit: '0.000001'
  currency: USD
--- a/api/core/model_runtime/model_providers/openai/llm/o1-2024-12-17.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/o1-2024-12-17.yaml
@ -0,0 +1,35 @@
+model: o1-2024-12-17
+label:
+  en_US: o1-2024-12-17
+model_type: llm
+features:
+  - multi-tool-call
+  - agent-thought
+  - stream-tool-call
+  - vision
+model_properties:
+  mode: chat
+  context_size: 200000
+parameter_rules:
+  - name: max_tokens
+    use_template: max_tokens
+    default: 50000
+    min: 1
+    max: 50000
+  - name: response_format
+    label:
+      zh_Hans: 回复格式
+      en_US: response_format
+    type: string
+    help:
+      zh_Hans: 指定模型必须输出的格式
+      en_US: specifying the format that the model must output
+    required: false
+    options:
+      - text
+      - json_object
+pricing:
+  input: '15.00'
+  output: '60.00'
+  unit: '0.000001'
+  currency: USD
--- a/api/core/model_runtime/model_providers/openai/llm/o1.yaml
+++ b/api/core/model_runtime/model_providers/openai/llm/o1.yaml
@ -0,0 +1,36 @@
+model: o1
+label:
+  zh_Hans: o1
+  en_US: o1
+model_type: llm
+features:
+  - multi-tool-call
+  - agent-thought
+  - stream-tool-call
+  - vision
+model_properties:
+  mode: chat
+  context_size: 200000
+parameter_rules:
+  - name: max_tokens
+    use_template: max_tokens
+    default: 50000
+    min: 1
+    max: 50000
+  - name: response_format
+    label:
+      zh_Hans: 回复格式
+      en_US: response_format
+    type: string
+    help:
+      zh_Hans: 指定模型必须输出的格式
+      en_US: specifying the format that the model must output
+    required: false
+    options:
+      - text
+      - json_object
+pricing:
+  input: '15.00'
+  output: '60.00'
+  unit: '0.000001'
+  currency: USD
--- a/api/core/rag/datasource/vdb/lindorm/lindorm_vector.py
+++ b/api/core/rag/datasource/vdb/lindorm/lindorm_vector.py
@ -49,10 +49,10 @@ class LindormVectorStoreConfig(BaseModel):


 class LindormVectorStore(BaseVector):
-    def __init__(self, collection_name: str, config: LindormVectorStoreConfig, **kwargs):
+    def __init__(self, collection_name: str, config: LindormVectorStoreConfig, using_ugc: bool, **kwargs):
        self._routing = None
        self._routing_field = None
-        if config.using_ugc:
+        if using_ugc:
            routing_value: str = kwargs.get("routing_value")
            if routing_value is None:
                raise ValueError("UGC index should init vector with valid 'routing_value' parameter value")
@ -64,7 +64,7 @@ class LindormVectorStore(BaseVector):
            super().__init__(collection_name.lower())
        self._client_config = config
        self._client = OpenSearch(**config.to_opensearch_params())
-        self._using_ugc = config.using_ugc
+        self._using_ugc = using_ugc
        self.kwargs = kwargs

    def get_type(self) -> str:
@ -467,12 +467,16 @@ class LindormVectorStoreFactory(AbstractVectorFactory):
        using_ugc = dify_config.USING_UGC_INDEX
        routing_value = None
        if dataset.index_struct:
-            if using_ugc:
+            # if an existed record's index_struct_dict doesn't contain using_ugc field,
+            # it actually stores in the normal index format
+            stored_in_ugc = dataset.index_struct_dict.get("using_ugc", False)
+            using_ugc = stored_in_ugc
+            if stored_in_ugc:
                dimension = dataset.index_struct_dict["dimension"]
                index_type = dataset.index_struct_dict["index_type"]
                distance_type = dataset.index_struct_dict["distance_type"]
-                index_name = f"{UGC_INDEX_PREFIX}_{dimension}_{index_type}_{distance_type}"
                routing_value = dataset.index_struct_dict["vector_store"]["class_prefix"]
+                index_name = f"{UGC_INDEX_PREFIX}_{dimension}_{index_type}_{distance_type}"
            else:
                index_name = dataset.index_struct_dict["vector_store"]["class_prefix"]
        else:
@ -487,6 +491,7 @@ class LindormVectorStoreFactory(AbstractVectorFactory):
                "index_type": index_type,
                "dimension": dimension,
                "distance_type": distance_type,
+                "using_ugc": using_ugc,
            }
            dataset.index_struct = json.dumps(index_struct_dict)
            if using_ugc:
@ -494,4 +499,4 @@ class LindormVectorStoreFactory(AbstractVectorFactory):
                routing_value = class_prefix
            else:
                index_name = class_prefix
-        return LindormVectorStore(index_name, lindorm_config, routing_value=routing_value)
+        return LindormVectorStore(index_name, lindorm_config, routing_value=routing_value, using_ugc=using_ugc)