gmakstutis/openrag
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

Merge pull request #55 from langflow-ai/lfx-openrag-update-flows

Browse source
This commit is contained in:
Edwin Jose 2025-10-02 18:23:36 -04:00 • committed by GitHub
commit 1628f84ef5
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
25 changed files with 6633 additions and 4966 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

BIN
.DS_Store vendored
View file

Binary file not shown.

2
Dockerfile.langflow
View file

@ -7,7 +7,7 @@ ENV RUSTFLAGS="--cfg reqwest_unstable"
# Accept build arguments for git repository and branch # Accept build arguments for git repository and branch
ARG GIT_REPO=https://github.com/langflow-ai/langflow.git ARG GIT_REPO=https://github.com/langflow-ai/langflow.git
ARG GIT_BRANCH=load_flows_autologin_false ARG GIT_BRANCH=test-openai-responses
WORKDIR /app WORKDIR /app

18
docker-compose.yml
View file

@ -40,9 +40,9 @@ services:
openrag-backend: openrag-backend:
image: phact/openrag-backend:${OPENRAG_VERSION:-latest} image: phact/openrag-backend:${OPENRAG_VERSION:-latest}
#build: # build:
#context: . # context: .
#dockerfile: Dockerfile.backend # dockerfile: Dockerfile.backend
container_name: openrag-backend container_name: openrag-backend
depends_on: depends_on:
- langflow - langflow
@ -77,9 +77,10 @@ services:
openrag-frontend: openrag-frontend:
image: phact/openrag-frontend:${OPENRAG_VERSION:-latest} image: phact/openrag-frontend:${OPENRAG_VERSION:-latest}
#build: # build:
#context: . # context: .
#dockerfile: Dockerfile.frontend # dockerfile: Dockerfile.frontend
# #dockerfile: Dockerfile.frontend
container_name: openrag-frontend container_name: openrag-frontend
depends_on: depends_on:
- openrag-backend - openrag-backend
@ -92,6 +93,9 @@ services:
volumes: volumes:
- ./flows:/app/flows:z - ./flows:/app/flows:z
image: phact/openrag-langflow:${LANGFLOW_VERSION:-latest} image: phact/openrag-langflow:${LANGFLOW_VERSION:-latest}
# build:
# context: .
# dockerfile: Dockerfile.langflow
container_name: langflow container_name: langflow
ports: ports:
- "7860:7860" - "7860:7860"
@ -99,7 +103,7 @@ services:
- OPENAI_API_KEY=${OPENAI_API_KEY} - OPENAI_API_KEY=${OPENAI_API_KEY}
- LANGFLOW_LOAD_FLOWS_PATH=/app/flows - LANGFLOW_LOAD_FLOWS_PATH=/app/flows
- LANGFLOW_SECRET_KEY=${LANGFLOW_SECRET_KEY} - LANGFLOW_SECRET_KEY=${LANGFLOW_SECRET_KEY}
- JWT="dummy" - JWT=None
- OWNER=None - OWNER=None
- OWNER_NAME=None - OWNER_NAME=None
- OWNER_EMAIL=None - OWNER_EMAIL=None

237
flows/components/ollama_embedding.json
View file

@ -1,28 +1,114 @@
{ {
"data": { "data": {
"id": "OllamaEmbeddings-4ah5Q",
"node": { "node": {
"base_classes": [ "template": {
"Embeddings" "_type": "Component",
], "base_url": {
"beta": false, "tool_mode": false,
"conditional_paths": [], "trace_as_input": true,
"custom_fields": {}, "trace_as_metadata": true,
"load_from_db": true,
"list": false,
"list_add_label": "Add More",
"required": true,
"placeholder": "",
"show": true,
"name": "base_url",
"value": "OLLAMA_BASE_URL",
"display_name": "Ollama Base URL",
"advanced": false,
"input_types": ["Message"],
"dynamic": false,
"info": "",
"title_case": false,
"type": "str",
"_input_type": "MessageTextInput"
},
"code": {
"type": "code",
"required": true,
"placeholder": "",
"list": false,
"show": true,
"multiline": true,
"value": "from typing import Any\nfrom urllib.parse import urljoin\n\nimport httpx\nfrom langchain_ollama import OllamaEmbeddings\n\nfrom lfx.base.models.model import LCModelComponent\nfrom lfx.base.models.ollama_constants import OLLAMA_EMBEDDING_MODELS, URL_LIST\nfrom lfx.field_typing import Embeddings\nfrom lfx.io import DropdownInput, MessageTextInput, Output\n\nHTTP_STATUS_OK = 200\n\n\nclass OllamaEmbeddingsComponent(LCModelComponent):\n display_name: str = \"Ollama Embeddings\"\n description: str = \"Generate embeddings using Ollama models.\"\n documentation = \"https://python.langchain.com/docs/integrations/text_embedding/ollama\"\n icon = \"Ollama\"\n name = \"OllamaEmbeddings\"\n\n inputs = [\n DropdownInput(\n name=\"model_name\",\n display_name=\"Ollama Model\",\n value=\"\",\n options=[],\n real_time_refresh=True,\n refresh_button=True,\n combobox=True,\n required=True,\n ),\n MessageTextInput(\n name=\"base_url\",\n display_name=\"Ollama Base URL\",\n value=\"\",\n required=True,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Embeddings\", name=\"embeddings\", method=\"build_embeddings\"),\n ]\n\n def build_embeddings(self) -> Embeddings:\n try:\n output = OllamaEmbeddings(model=self.model_name, base_url=self.base_url)\n except Exception as e:\n msg = (\n \"Unable to connect to the Ollama API. \",\n \"Please verify the base URL, ensure the relevant Ollama model is pulled, and try again.\",\n )\n raise ValueError(msg) from e\n return output\n\n async def update_build_config(self, build_config: dict, field_value: Any, field_name: str | None = None):\n if field_name in {\"base_url\", \"model_name\"} and not await self.is_valid_ollama_url(field_value):\n # Check if any URL in the list is valid\n valid_url = \"\"\n for url in URL_LIST:\n if await self.is_valid_ollama_url(url):\n valid_url = url\n break\n build_config[\"base_url\"][\"value\"] = valid_url\n if field_name in {\"model_name\", \"base_url\", \"tool_model_enabled\"}:\n if await self.is_valid_ollama_url(self.base_url):\n build_config[\"model_name\"][\"options\"] = await self.get_model(self.base_url)\n elif await self.is_valid_ollama_url(build_config[\"base_url\"].get(\"value\", \"\")):\n build_config[\"model_name\"][\"options\"] = await self.get_model(build_config[\"base_url\"].get(\"value\", \"\"))\n else:\n build_config[\"model_name\"][\"options\"] = []\n\n return build_config\n\n async def get_model(self, base_url_value: str) -> list[str]:\n \"\"\"Get the model names from Ollama.\"\"\"\n model_ids = []\n try:\n url = urljoin(base_url_value, \"/api/tags\")\n async with httpx.AsyncClient() as client:\n response = await client.get(url)\n response.raise_for_status()\n data = response.json()\n\n model_ids = [model[\"name\"] for model in data.get(\"models\", [])]\n # this to ensure that not embedding models are included.\n # not even the base models since models can have 1b 2b etc\n # handles cases when embeddings models have tags like :latest - etc.\n model_ids = [\n model\n for model in model_ids\n if any(model.startswith(f\"{embedding_model}\") for embedding_model in OLLAMA_EMBEDDING_MODELS)\n ]\n\n except (ImportError, ValueError, httpx.RequestError) as e:\n msg = \"Could not get model names from Ollama.\"\n raise ValueError(msg) from e\n\n return model_ids\n\n async def is_valid_ollama_url(self, url: str) -> bool:\n try:\n async with httpx.AsyncClient() as client:\n return (await client.get(f\"{url}/api/tags\")).status_code == HTTP_STATUS_OK\n except httpx.RequestError:\n return False\n",
"fileTypes": [],
"file_path": "",
"password": false,
"name": "code",
"advanced": true,
"dynamic": true,
"info": "",
"load_from_db": false,
"title_case": false
},
"model_name": {
"tool_mode": false,
"trace_as_metadata": true,
"options": ["nomic-embed-text:latest", "all-minilm:latest"],
"options_metadata": [],
"combobox": true,
"dialog_inputs": {},
"toggle": false,
"required": true,
"placeholder": "",
"show": true,
"name": "model_name",
"value": "",
"display_name": "Ollama Model",
"advanced": false,
"dynamic": false,
"info": "",
"real_time_refresh": true,
"refresh_button": true,
"title_case": false,
"external_options": {},
"type": "str",
"_input_type": "DropdownInput"
}
},
"description": "Generate embeddings using Ollama models.", "description": "Generate embeddings using Ollama models.",
"icon": "Ollama",
"base_classes": ["Embeddings"],
"display_name": "Ollama Embeddings", "display_name": "Ollama Embeddings",
"documentation": "https://python.langchain.com/docs/integrations/text_embedding/ollama", "documentation": "https://python.langchain.com/docs/integrations/text_embedding/ollama",
"edited": false, "minimized": false,
"field_order": [ "custom_fields": {},
"model_name", "output_types": [],
"base_url" "pinned": false,
], "conditional_paths": [],
"frozen": false, "frozen": false,
"icon": "Ollama", "outputs": [
"last_updated": "2025-09-22T20:18:27.128Z", {
"types": ["Embeddings"],
"selected": "Embeddings",
"name": "embeddings",
"display_name": "Embeddings",
"method": "build_embeddings",
"value": "__UNDEFINED__",
"cache": true,
"required_inputs": null,
"allows_loop": false,
"group_outputs": false,
"options": null,
"tool_mode": true
}
],
"field_order": ["model_name", "base_url"],
"beta": false,
"legacy": false, "legacy": false,
"edited": false,
"metadata": { "metadata": {
"code_hash": "0db0f99e91e9", "keywords": [
"model",
"llm",
"language model",
"large language model"
],
"module": "lfx.components.ollama.ollama_embeddings.OllamaEmbeddingsComponent",
"code_hash": "c41821735548",
"dependencies": { "dependencies": {
"total_dependencies": 3,
"dependencies": [ "dependencies": [
{ {
"name": "httpx", "name": "httpx",
@ -33,125 +119,24 @@
"version": "0.2.1" "version": "0.2.1"
}, },
{ {
"name": "langflow", "name": "lfx",
"version": null "version": null
} }
], ]
"total_dependencies": 3
},
"keywords": [
"model",
"llm",
"language model",
"large language model"
],
"module": "langflow.components.ollama.ollama_embeddings.OllamaEmbeddingsComponent"
},
"minimized": false,
"output_types": [],
"outputs": [
{
"allows_loop": false,
"cache": true,
"display_name": "Embeddings",
"group_outputs": false,
"method": "build_embeddings",
"name": "embeddings",
"options": null,
"required_inputs": null,
"selected": "Embeddings",
"tool_mode": true,
"types": [
"Embeddings"
],
"value": "__UNDEFINED__"
}
],
"pinned": false,
"template": {
"_type": "Component",
"base_url": {
"_input_type": "MessageTextInput",
"advanced": false,
"display_name": "Ollama Base URL",
"dynamic": false,
"info": "",
"input_types": [
"Message"
],
"list": false,
"list_add_label": "Add More",
"load_from_db": true,
"name": "base_url",
"placeholder": "",
"required": true,
"show": true,
"title_case": false,
"tool_mode": false,
"trace_as_input": true,
"trace_as_metadata": true,
"type": "str",
"value": "OLLAMA_BASE_URL"
},
"code": {
"advanced": true,
"dynamic": true,
"fileTypes": [],
"file_path": "",
"info": "",
"list": false,
"load_from_db": false,
"multiline": true,
"name": "code",
"password": false,
"placeholder": "",
"required": true,
"show": true,
"title_case": false,
"type": "code",
"value": "from typing import Any\nfrom urllib.parse import urljoin\n\nimport httpx\nfrom langchain_ollama import OllamaEmbeddings\n\nfrom langflow.base.models.model import LCModelComponent\nfrom langflow.base.models.ollama_constants import OLLAMA_EMBEDDING_MODELS, URL_LIST\nfrom langflow.field_typing import Embeddings\nfrom langflow.io import DropdownInput, MessageTextInput, Output\n\nHTTP_STATUS_OK = 200\n\n\nclass OllamaEmbeddingsComponent(LCModelComponent):\n display_name: str = \"Ollama Embeddings\"\n description: str = \"Generate embeddings using Ollama models.\"\n documentation = \"https://python.langchain.com/docs/integrations/text_embedding/ollama\"\n icon = \"Ollama\"\n name = \"OllamaEmbeddings\"\n\n inputs = [\n DropdownInput(\n name=\"model_name\",\n display_name=\"Ollama Model\",\n value=\"\",\n options=[],\n real_time_refresh=True,\n refresh_button=True,\n combobox=True,\n required=True,\n ),\n MessageTextInput(\n name=\"base_url\",\n display_name=\"Ollama Base URL\",\n value=\"\",\n required=True,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Embeddings\", name=\"embeddings\", method=\"build_embeddings\"),\n ]\n\n def build_embeddings(self) -> Embeddings:\n try:\n output = OllamaEmbeddings(model=self.model_name, base_url=self.base_url)\n except Exception as e:\n msg = (\n \"Unable to connect to the Ollama API. \",\n \"Please verify the base URL, ensure the relevant Ollama model is pulled, and try again.\",\n )\n raise ValueError(msg) from e\n return output\n\n async def update_build_config(self, build_config: dict, field_value: Any, field_name: str | None = None):\n if field_name in {\"base_url\", \"model_name\"} and not await self.is_valid_ollama_url(field_value):\n # Check if any URL in the list is valid\n valid_url = \"\"\n for url in URL_LIST:\n if await self.is_valid_ollama_url(url):\n valid_url = url\n break\n build_config[\"base_url\"][\"value\"] = valid_url\n if field_name in {\"model_name\", \"base_url\", \"tool_model_enabled\"}:\n if await self.is_valid_ollama_url(self.base_url):\n build_config[\"model_name\"][\"options\"] = await self.get_model(self.base_url)\n elif await self.is_valid_ollama_url(build_config[\"base_url\"].get(\"value\", \"\")):\n build_config[\"model_name\"][\"options\"] = await self.get_model(build_config[\"base_url\"].get(\"value\", \"\"))\n else:\n build_config[\"model_name\"][\"options\"] = []\n\n return build_config\n\n async def get_model(self, base_url_value: str) -> list[str]:\n \"\"\"Get the model names from Ollama.\"\"\"\n model_ids = []\n try:\n url = urljoin(base_url_value, \"/api/tags\")\n async with httpx.AsyncClient() as client:\n response = await client.get(url)\n response.raise_for_status()\n data = response.json()\n\n model_ids = [model[\"name\"] for model in data.get(\"models\", [])]\n # this to ensure that not embedding models are included.\n # not even the base models since models can have 1b 2b etc\n # handles cases when embeddings models have tags like :latest - etc.\n model_ids = [\n model\n for model in model_ids\n if any(model.startswith(f\"{embedding_model}\") for embedding_model in OLLAMA_EMBEDDING_MODELS)\n ]\n\n except (ImportError, ValueError, httpx.RequestError) as e:\n msg = \"Could not get model names from Ollama.\"\n raise ValueError(msg) from e\n\n return model_ids\n\n async def is_valid_ollama_url(self, url: str) -> bool:\n try:\n async with httpx.AsyncClient() as client:\n return (await client.get(f\"{url}/api/tags\")).status_code == HTTP_STATUS_OK\n except httpx.RequestError:\n return False\n"
},
"model_name": {
"_input_type": "DropdownInput",
"advanced": false,
"combobox": true,
"dialog_inputs": {},
"display_name": "Ollama Model",
"dynamic": false,
"info": "",
"name": "model_name",
"options": [
"all-minilm:latest"
],
"options_metadata": [],
"placeholder": "",
"real_time_refresh": true,
"refresh_button": true,
"required": true,
"show": true,
"title_case": false,
"toggle": false,
"tool_mode": false,
"trace_as_metadata": true,
"type": "str",
"value": "all-minilm:latest"
} }
}, },
"tool_mode": false "tool_mode": false,
"last_updated": "2025-09-29T18:40:10.242Z",
"official": false
}, },
"showNode": true, "showNode": true,
"type": "OllamaEmbeddings" "type": "OllamaEmbeddings",
}, "id": "OllamaEmbeddings-vnNn8"
"dragging": false,
"id": "OllamaEmbeddings-4ah5Q",
"measured": {
"height": 286,
"width": 320
}, },
"id": "OllamaEmbeddings-vnNn8",
"position": { "position": {
"x": 282.29416840859585, "x": 0,
"y": 279.4218065717267 "y": 0
}, },
"selected": false,
"type": "genericNode" "type": "genericNode"
} }

1261
flows/components/ollama_llm.json
View file

File diff suppressed because one or more lines are too long

1354
flows/components/ollama_llm_text.json
View file

File diff suppressed because one or more lines are too long

401
flows/components/watsonx_embedding.json
View file

File diff suppressed because one or more lines are too long

959
flows/components/watsonx_llm.json
View file

File diff suppressed because one or more lines are too long

1058
flows/components/watsonx_llm_text.json
View file

File diff suppressed because one or more lines are too long

1108
flows/ingestion_flow.json
View file

File diff suppressed because one or more lines are too long

706
flows/openrag_agent.json
View file

File diff suppressed because one or more lines are too long

1014
flows/openrag_ingest_docling.json
View file

File diff suppressed because one or more lines are too long

570
flows/openrag_nudges.json
View file

File diff suppressed because one or more lines are too long

46
frontend/components/ui/switch.tsx
View file

@ -1,29 +1,29 @@
"use client" "use client";
import * as React from "react" import * as SwitchPrimitives from "@radix-ui/react-switch";
import * as SwitchPrimitives from "@radix-ui/react-switch" import * as React from "react";
import { cn } from "@/lib/utils" import { cn } from "@/lib/utils";
const Switch = React.forwardRef< const Switch = React.forwardRef<
React.ElementRef<typeof SwitchPrimitives.Root>, React.ElementRef<typeof SwitchPrimitives.Root>,
React.ComponentPropsWithoutRef<typeof SwitchPrimitives.Root> React.ComponentPropsWithoutRef<typeof SwitchPrimitives.Root>
>(({ className, ...props }, ref) => ( >(({ className, ...props }, ref) => (
<SwitchPrimitives.Root <SwitchPrimitives.Root
className={cn( className={cn(
"peer inline-flex h-6 w-11 shrink-0 cursor-pointer items-center rounded-full border-2 border-transparent transition-colors focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2 focus-visible:ring-offset-background disabled:cursor-not-allowed disabled:opacity-50 data-[state=checked]:bg-primary data-[state=unchecked]:bg-muted", "peer inline-flex h-6 w-11 shrink-0 cursor-pointer items-center rounded-full border-2 border-transparent transition-colors focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2 focus-visible:ring-offset-background disabled:cursor-not-allowed disabled:opacity-50 data-[state=checked]:bg-primary data-[state=unchecked]:bg-muted",
className className,
)} )}
{...props} {...props}
ref={ref} ref={ref}
> >
<SwitchPrimitives.Thumb <SwitchPrimitives.Thumb
className={cn( className={cn(
"pointer-events-none block h-5 w-5 rounded-full bg-background shadow-lg ring-0 transition-transform data-[state=checked]:translate-x-5 data-[state=unchecked]:translate-x-0 data-[state=unchecked]:bg-primary" "pointer-events-none block h-5 w-5 rounded-full bg-background shadow-lg ring-0 transition-transform data-[state=checked]:translate-x-5 data-[state=unchecked]:translate-x-0 data-[state=unchecked]:bg-primary",
)} )}
/> />
</SwitchPrimitives.Root> </SwitchPrimitives.Root>
)) ));
Switch.displayName = SwitchPrimitives.Root.displayName Switch.displayName = SwitchPrimitives.Root.displayName;
export { Switch } export { Switch };

156
frontend/src/app/settings/helpers/model-helpers.tsx
View file

@ -1,87 +1,97 @@
import OpenAILogo from "@/components/logo/openai-logo";
import OllamaLogo from "@/components/logo/ollama-logo";
import IBMLogo from "@/components/logo/ibm-logo"; import IBMLogo from "@/components/logo/ibm-logo";
import OllamaLogo from "@/components/logo/ollama-logo";
import OpenAILogo from "@/components/logo/openai-logo";
export type ModelProvider = 'openai' | 'ollama' | 'ibm'; export type ModelProvider = "openai" | "ollama" | "watsonx";
export interface ModelOption { export interface ModelOption {
value: string; value: string;
label: string; label: string;
} }
// Helper function to get model logo based on provider or model name // Helper function to get model logo based on provider or model name
export function getModelLogo(modelValue: string, provider?: ModelProvider) { export function getModelLogo(modelValue: string, provider?: ModelProvider) {
// First check by provider // First check by provider
if (provider === 'openai') { if (provider === "openai") {
return <OpenAILogo className="w-4 h-4" />; return <OpenAILogo className="w-4 h-4" />;
} else if (provider === 'ollama') { } else if (provider === "ollama") {
return <OllamaLogo className="w-4 h-4" />; return <OllamaLogo className="w-4 h-4" />;
} else if (provider === 'ibm') { } else if (provider === "watsonx") {
return <IBMLogo className="w-4 h-4" />; return <IBMLogo className="w-4 h-4" />;
} }
// Fallback to model name analysis // Fallback to model name analysis
if (modelValue.includes('gpt') || modelValue.includes('text-embedding')) { if (modelValue.includes("gpt") || modelValue.includes("text-embedding")) {
return <OpenAILogo className="w-4 h-4" />; return <OpenAILogo className="w-4 h-4" />;
} else if (modelValue.includes('llama') || modelValue.includes('ollama')) { } else if (modelValue.includes("llama") || modelValue.includes("ollama")) {
return <OllamaLogo className="w-4 h-4" />; return <OllamaLogo className="w-4 h-4" />;
} else if (modelValue.includes('granite') || modelValue.includes('slate') || modelValue.includes('ibm')) { } else if (
return <IBMLogo className="w-4 h-4" />; modelValue.includes("granite") ||
} modelValue.includes("slate") ||
modelValue.includes("ibm")
) {
return <IBMLogo className="w-4 h-4" />;
}
return <OpenAILogo className="w-4 h-4" />; // Default to OpenAI logo return <OpenAILogo className="w-4 h-4" />; // Default to OpenAI logo
} }
// Helper function to get fallback models by provider // Helper function to get fallback models by provider
export function getFallbackModels(provider: ModelProvider) { export function getFallbackModels(provider: ModelProvider) {
switch (provider) { switch (provider) {
case 'openai': case "openai":
return { return {
language: [ language: [
{ value: 'gpt-4', label: 'GPT-4' }, { value: "gpt-4", label: "GPT-4" },
{ value: 'gpt-4-turbo', label: 'GPT-4 Turbo' }, { value: "gpt-4-turbo", label: "GPT-4 Turbo" },
{ value: 'gpt-3.5-turbo', label: 'GPT-3.5 Turbo' }, { value: "gpt-3.5-turbo", label: "GPT-3.5 Turbo" },
], ],
embedding: [ embedding: [
{ value: 'text-embedding-ada-002', label: 'text-embedding-ada-002' }, { value: "text-embedding-ada-002", label: "text-embedding-ada-002" },
{ value: 'text-embedding-3-small', label: 'text-embedding-3-small' }, { value: "text-embedding-3-small", label: "text-embedding-3-small" },
{ value: 'text-embedding-3-large', label: 'text-embedding-3-large' }, { value: "text-embedding-3-large", label: "text-embedding-3-large" },
], ],
}; };
case 'ollama': case "ollama":
return { return {
language: [ language: [
{ value: 'llama2', label: 'Llama 2' }, { value: "llama2", label: "Llama 2" },
{ value: 'llama2:13b', label: 'Llama 2 13B' }, { value: "llama2:13b", label: "Llama 2 13B" },
{ value: 'codellama', label: 'Code Llama' }, { value: "codellama", label: "Code Llama" },
], ],
embedding: [ embedding: [
{ value: 'mxbai-embed-large', label: 'MxBai Embed Large' }, { value: "mxbai-embed-large", label: "MxBai Embed Large" },
{ value: 'nomic-embed-text', label: 'Nomic Embed Text' }, { value: "nomic-embed-text", label: "Nomic Embed Text" },
], ],
}; };
case 'ibm': case "watsonx":
return { return {
language: [ language: [
{ value: 'meta-llama/llama-3-1-70b-instruct', label: 'Llama 3.1 70B Instruct' }, {
{ value: 'ibm/granite-13b-chat-v2', label: 'Granite 13B Chat v2' }, value: "meta-llama/llama-3-1-70b-instruct",
], label: "Llama 3.1 70B Instruct",
embedding: [ },
{ value: 'ibm/slate-125m-english-rtrvr', label: 'Slate 125M English Retriever' }, { value: "ibm/granite-13b-chat-v2", label: "Granite 13B Chat v2" },
], ],
}; embedding: [
default: {
return { value: "ibm/slate-125m-english-rtrvr",
language: [ label: "Slate 125M English Retriever",
{ value: 'gpt-4', label: 'GPT-4' }, },
{ value: 'gpt-4-turbo', label: 'GPT-4 Turbo' }, ],
{ value: 'gpt-3.5-turbo', label: 'GPT-3.5 Turbo' }, };
], default:
embedding: [ return {
{ value: 'text-embedding-ada-002', label: 'text-embedding-ada-002' }, language: [
{ value: 'text-embedding-3-small', label: 'text-embedding-3-small' }, { value: "gpt-4", label: "GPT-4" },
{ value: 'text-embedding-3-large', label: 'text-embedding-3-large' }, { value: "gpt-4-turbo", label: "GPT-4 Turbo" },
], { value: "gpt-3.5-turbo", label: "GPT-3.5 Turbo" },
}; ],
} embedding: [
} { value: "text-embedding-ada-002", label: "text-embedding-ada-002" },
{ value: "text-embedding-3-small", label: "text-embedding-3-small" },
{ value: "text-embedding-3-large", label: "text-embedding-3-large" },
],
};
}
}

2341
frontend/src/app/settings/page.tsx
View file

File diff suppressed because it is too large Load diff

14
src/api/langflow_files.py
View file

@ -70,22 +70,22 @@ async def run_ingestion(
if settings: if settings:
logger.debug("Applying ingestion settings", settings=settings) logger.debug("Applying ingestion settings", settings=settings)
# Split Text component tweaks (SplitText-QIKhg) # Split Text component tweaks (SplitText-PC36h)
if ( if (
settings.get("chunkSize") settings.get("chunkSize")
or settings.get("chunkOverlap") or settings.get("chunkOverlap")
or settings.get("separator") or settings.get("separator")
): ):
if "SplitText-QIKhg" not in tweaks: if "SplitText-PC36h" not in tweaks:
tweaks["SplitText-QIKhg"] = {} tweaks["SplitText-PC36h"] = {}
if settings.get("chunkSize"): if settings.get("chunkSize"):
tweaks["SplitText-QIKhg"]["chunk_size"] = settings["chunkSize"] tweaks["SplitText-PC36h"]["chunk_size"] = settings["chunkSize"]
if settings.get("chunkOverlap"): if settings.get("chunkOverlap"):
tweaks["SplitText-QIKhg"]["chunk_overlap"] = settings[ tweaks["SplitText-PC36h"]["chunk_overlap"] = settings[
"chunkOverlap" "chunkOverlap"
] ]
if settings.get("separator"): if settings.get("separator"):
tweaks["SplitText-QIKhg"]["separator"] = settings["separator"] tweaks["SplitText-PC36h"]["separator"] = settings["separator"]
# OpenAI Embeddings component tweaks (OpenAIEmbeddings-joRJ6) # OpenAI Embeddings component tweaks (OpenAIEmbeddings-joRJ6)
if settings.get("embeddingModel"): if settings.get("embeddingModel"):
@ -275,4 +275,4 @@ async def delete_user_files(
status_code=status, status_code=status,
) )
except Exception as e: except Exception as e:
return JSONResponse({"error": str(e)}, status_code=500) return JSONResponse({"error": str(e)}, status_code=500)

8
src/api/router.py
View file

@ -140,6 +140,14 @@ async def langflow_upload_ingest_task(
) )
# Create langflow upload task # Create langflow upload task
print(f"tweaks: {tweaks}")
print(f"settings: {settings}")
print(f"jwt_token: {jwt_token}")
print(f"user_name: {user_name}")
print(f"user_email: {user_email}")
print(f"session_id: {session_id}")
print(f"delete_after_ingest: {delete_after_ingest}")
print(f"temp_file_paths: {temp_file_paths}")
task_id = await task_service.create_langflow_upload_task( task_id = await task_service.create_langflow_upload_task(
user_id=user_id, user_id=user_id,
file_paths=temp_file_paths, file_paths=temp_file_paths,

9
src/api/settings.py
View file

@ -9,7 +9,6 @@ from config.settings import (
LANGFLOW_CHAT_FLOW_ID, LANGFLOW_CHAT_FLOW_ID,
LANGFLOW_INGEST_FLOW_ID, LANGFLOW_INGEST_FLOW_ID,
LANGFLOW_PUBLIC_URL, LANGFLOW_PUBLIC_URL,
DOCLING_COMPONENT_ID,
LOCALHOST_URL, LOCALHOST_URL,
clients, clients,
get_openrag_config, get_openrag_config,
@ -206,7 +205,7 @@ async def update_settings(request, session_manager):
# Also update the chat flow with the new model # Also update the chat flow with the new model
try: try:
flows_service = _get_flows_service() flows_service = _get_flows_service()
await flows_service.update_chat_flow_model(body["llm_model"]) await flows_service.update_chat_flow_model(body["llm_model"], current_config.provider.model_provider.lower())
logger.info( logger.info(
f"Successfully updated chat flow model to '{body['llm_model']}'" f"Successfully updated chat flow model to '{body['llm_model']}'"
) )
@ -223,7 +222,8 @@ async def update_settings(request, session_manager):
try: try:
flows_service = _get_flows_service() flows_service = _get_flows_service()
await flows_service.update_chat_flow_system_prompt( await flows_service.update_chat_flow_system_prompt(
body["system_prompt"] body["system_prompt"],
current_config.provider.model_provider.lower()
) )
logger.info(f"Successfully updated chat flow system prompt") logger.info(f"Successfully updated chat flow system prompt")
except Exception as e: except Exception as e:
@ -248,7 +248,8 @@ async def update_settings(request, session_manager):
try: try:
flows_service = _get_flows_service() flows_service = _get_flows_service()
await flows_service.update_ingest_flow_embedding_model( await flows_service.update_ingest_flow_embedding_model(
body["embedding_model"].strip() body["embedding_model"].strip(),
current_config.provider.model_provider.lower()
) )
logger.info( logger.info(
f"Successfully updated ingest flow embedding model to '{body['embedding_model'].strip()}'" f"Successfully updated ingest flow embedding model to '{body['embedding_model'].strip()}'"

34
src/config/settings.py
View file

@ -543,38 +543,28 @@ OLLAMA_EMBEDDING_COMPONENT_PATH = os.getenv(
# Component IDs in flows # Component IDs in flows
OPENAI_EMBEDDING_COMPONENT_ID = os.getenv( OPENAI_EMBEDDING_COMPONENT_DISPLAY_NAME = os.getenv(
"OPENAI_EMBEDDING_COMPONENT_ID", "EmbeddingModel-eZ6bT" "OPENAI_EMBEDDING_COMPONENT_DISPLAY_NAME", "Embedding Model"
) )
OPENAI_LLM_COMPONENT_ID = os.getenv( OPENAI_LLM_COMPONENT_DISPLAY_NAME = os.getenv(
"OPENAI_LLM_COMPONENT_ID", "LanguageModelComponent-0YME7" "OPENAI_LLM_COMPONENT_DISPLAY_NAME", "Language Model"
)
OPENAI_LLM_TEXT_COMPONENT_ID = os.getenv(
"OPENAI_LLM_TEXT_COMPONENT_ID", "LanguageModelComponent-NSTA6"
) )
# Provider-specific component IDs # Provider-specific component IDs
WATSONX_EMBEDDING_COMPONENT_ID = os.getenv( WATSONX_EMBEDDING_COMPONENT_DISPLAY_NAME = os.getenv(
"WATSONX_EMBEDDING_COMPONENT_ID", "WatsonxEmbeddingsComponent-pJfXI" "WATSONX_EMBEDDING_COMPONENT_DISPLAY_NAME", "IBM watsonx.ai Embeddings"
) )
WATSONX_LLM_COMPONENT_ID = os.getenv( WATSONX_LLM_COMPONENT_DISPLAY_NAME = os.getenv(
"WATSONX_LLM_COMPONENT_ID", "IBMwatsonxModel-jA4Nw" "WATSONX_LLM_COMPONENT_DISPLAY_NAME", "IBM watsonx.ai"
)
WATSONX_LLM_TEXT_COMPONENT_ID = os.getenv(
"WATSONX_LLM_TEXT_COMPONENT_ID", "IBMwatsonxModel-18kmA"
) )
OLLAMA_EMBEDDING_COMPONENT_DISPLAY_NAME = os.getenv(
OLLAMA_EMBEDDING_COMPONENT_ID = os.getenv( "OLLAMA_EMBEDDING_COMPONENT_DISPLAY_NAME", "Ollama Model"
"OLLAMA_EMBEDDING_COMPONENT_ID", "OllamaEmbeddings-4ah5Q"
)
OLLAMA_LLM_COMPONENT_ID = os.getenv("OLLAMA_LLM_COMPONENT_ID", "OllamaModel-eCsJx")
OLLAMA_LLM_TEXT_COMPONENT_ID = os.getenv(
"OLLAMA_LLM_TEXT_COMPONENT_ID", "OllamaModel-XDGqZ"
) )
OLLAMA_LLM_COMPONENT_DISPLAY_NAME = os.getenv("OLLAMA_LLM_COMPONENT_DISPLAY_NAME", "Ollama")
# Docling component ID for ingest flow # Docling component ID for ingest flow
DOCLING_COMPONENT_ID = os.getenv("DOCLING_COMPONENT_ID", "DoclingRemote-78KoX") DOCLING_COMPONENT_DISPLAY_NAME = os.getenv("DOCLING_COMPONENT_DISPLAY_NAME", "Docling Serve")
LOCALHOST_URL = get_container_host() or "localhost" LOCALHOST_URL = get_container_host() or "localhost"

6
src/main.py
View file

@ -382,6 +382,10 @@ async def _ingest_default_documents_langflow(services, file_paths):
settings=None, # Use default ingestion settings settings=None, # Use default ingestion settings
jwt_token=effective_jwt, # Use JWT token (anonymous if needed) jwt_token=effective_jwt, # Use JWT token (anonymous if needed)
delete_after_ingest=True, # Clean up after ingestion delete_after_ingest=True, # Clean up after ingestion
owner=None,
owner_name=anonymous_user.name,
owner_email=anonymous_user.email,
connector_type="system_default",
) )
logger.info( logger.info(
@ -1189,4 +1193,4 @@ if __name__ == "__main__":
host="0.0.0.0", host="0.0.0.0",
port=8000, port=8000,
reload=False, # Disable reload since we're running from main reload=False, # Disable reload since we're running from main
) )

9
src/models/processors.py
View file

@ -625,7 +625,12 @@ class LangflowFileProcessor(TaskProcessor):
tweaks=final_tweaks, tweaks=final_tweaks,
settings=self.settings, settings=self.settings,
jwt_token=effective_jwt, jwt_token=effective_jwt,
delete_after_ingest=self.delete_after_ingest delete_after_ingest=self.delete_after_ingest,
owner=self.owner_user_id,
owner_name=self.owner_name,
owner_email=self.owner_email,
connector_type="local",
) )
# Update task with success # Update task with success
@ -640,4 +645,4 @@ class LangflowFileProcessor(TaskProcessor):
file_task.error_message = str(e) file_task.error_message = str(e)
file_task.updated_at = time.time() file_task.updated_at = time.time()
upload_task.failed_files += 1 upload_task.failed_files += 1
raise raise

17
src/services/chat_service.py
View file

@ -110,15 +110,14 @@ class ChatService:
filter_expression["score_threshold"] = score_threshold filter_expression["score_threshold"] = score_threshold
# Pass the complete filter expression as a single header to Langflow (only if we have something to send) # Pass the complete filter expression as a single header to Langflow (only if we have something to send)
if filter_expression: logger.info(
logger.info( "Sending OpenRAG query filter to Langflow",
"Sending OpenRAG query filter to Langflow", filter_expression=filter_expression,
filter_expression=filter_expression, )
) extra_headers["X-LANGFLOW-GLOBAL-VAR-OPENRAG-QUERY-FILTER"] = json.dumps(
extra_headers["X-LANGFLOW-GLOBAL-VAR-OPENRAG-QUERY-FILTER"] = json.dumps( filter_expression
filter_expression )
) logger.info(f"[LF] Extra headers {extra_headers}")
# Ensure the Langflow client exists; try lazy init if needed # Ensure the Langflow client exists; try lazy init if needed
langflow_client = await clients.ensure_langflow_client() langflow_client = await clients.ensure_langflow_client()
if not langflow_client: if not langflow_client:

233
src/services/flows_service.py
View file

@ -1,26 +1,22 @@
import asyncio
from config.settings import ( from config.settings import (
DISABLE_INGEST_WITH_LANGFLOW, DISABLE_INGEST_WITH_LANGFLOW,
NUDGES_FLOW_ID, NUDGES_FLOW_ID,
LANGFLOW_URL, LANGFLOW_URL,
LANGFLOW_CHAT_FLOW_ID, LANGFLOW_CHAT_FLOW_ID,
LANGFLOW_INGEST_FLOW_ID, LANGFLOW_INGEST_FLOW_ID,
OLLAMA_LLM_TEXT_COMPONENT_ID,
OLLAMA_LLM_TEXT_COMPONENT_PATH, OLLAMA_LLM_TEXT_COMPONENT_PATH,
OPENAI_EMBEDDING_COMPONENT_ID, OPENAI_EMBEDDING_COMPONENT_DISPLAY_NAME,
OPENAI_LLM_COMPONENT_ID, OPENAI_LLM_COMPONENT_DISPLAY_NAME,
OPENAI_LLM_TEXT_COMPONENT_ID,
WATSONX_LLM_TEXT_COMPONENT_ID,
WATSONX_LLM_TEXT_COMPONENT_PATH, WATSONX_LLM_TEXT_COMPONENT_PATH,
clients, clients,
WATSONX_LLM_COMPONENT_PATH, WATSONX_LLM_COMPONENT_PATH,
WATSONX_EMBEDDING_COMPONENT_PATH, WATSONX_EMBEDDING_COMPONENT_PATH,
OLLAMA_LLM_COMPONENT_PATH, OLLAMA_LLM_COMPONENT_PATH,
OLLAMA_EMBEDDING_COMPONENT_PATH, OLLAMA_EMBEDDING_COMPONENT_PATH,
WATSONX_EMBEDDING_COMPONENT_ID, WATSONX_EMBEDDING_COMPONENT_DISPLAY_NAME,
WATSONX_LLM_COMPONENT_ID, WATSONX_LLM_COMPONENT_DISPLAY_NAME,
OLLAMA_EMBEDDING_COMPONENT_ID, OLLAMA_EMBEDDING_COMPONENT_DISPLAY_NAME,
OLLAMA_LLM_COMPONENT_ID, OLLAMA_LLM_COMPONENT_DISPLAY_NAME,
get_openrag_config, get_openrag_config,
) )
import json import json
@ -277,23 +273,23 @@ class FlowsService:
{ {
"name": "nudges", "name": "nudges",
"flow_id": NUDGES_FLOW_ID, "flow_id": NUDGES_FLOW_ID,
"embedding_id": OPENAI_EMBEDDING_COMPONENT_ID, "embedding_name": OPENAI_EMBEDDING_COMPONENT_DISPLAY_NAME,
"llm_id": OPENAI_LLM_COMPONENT_ID, "llm_text_name": OPENAI_LLM_COMPONENT_DISPLAY_NAME,
"llm_text_id": OPENAI_LLM_TEXT_COMPONENT_ID, "llm_name": None,
}, },
{ {
"name": "retrieval", "name": "retrieval",
"flow_id": LANGFLOW_CHAT_FLOW_ID, "flow_id": LANGFLOW_CHAT_FLOW_ID,
"embedding_id": OPENAI_EMBEDDING_COMPONENT_ID, "embedding_name": OPENAI_EMBEDDING_COMPONENT_DISPLAY_NAME,
"llm_id": OPENAI_LLM_COMPONENT_ID, "llm_name": OPENAI_LLM_COMPONENT_DISPLAY_NAME,
"llm_text_id": None, "llm_text_name": None,
}, },
{ {
"name": "ingest", "name": "ingest",
"flow_id": LANGFLOW_INGEST_FLOW_ID, "flow_id": LANGFLOW_INGEST_FLOW_ID,
"embedding_id": OPENAI_EMBEDDING_COMPONENT_ID, "embedding_name": OPENAI_EMBEDDING_COMPONENT_DISPLAY_NAME,
"llm_id": None, # Ingestion flow might not have LLM "llm_name": None, # Ingestion flow might not have LLM
"llm_text_id": None, # Ingestion flow might not have LLM Text "llm_text_name": None,
}, },
] ]
@ -389,9 +385,9 @@ class FlowsService:
"""Update components in a specific flow""" """Update components in a specific flow"""
flow_name = config["name"] flow_name = config["name"]
flow_id = config["flow_id"] flow_id = config["flow_id"]
old_embedding_id = config["embedding_id"] old_embedding_name = config["embedding_name"]
old_llm_id = config["llm_id"] old_llm_name = config["llm_name"]
old_llm_text_id = config["llm_text_id"] old_llm_text_name = config["llm_text_name"]
# Extract IDs from templates # Extract IDs from templates
new_llm_id = llm_template["data"]["id"] new_llm_id = llm_template["data"]["id"]
new_embedding_id = embedding_template["data"]["id"] new_embedding_id = embedding_template["data"]["id"]
@ -411,7 +407,7 @@ class FlowsService:
# Replace embedding component # Replace embedding component
if not DISABLE_INGEST_WITH_LANGFLOW: if not DISABLE_INGEST_WITH_LANGFLOW:
embedding_node = self._find_node_by_id(flow_data, old_embedding_id) embedding_node, _ = self._find_node_in_flow(flow_data, display_name=old_embedding_name)
if embedding_node: if embedding_node:
# Preserve position # Preserve position
original_position = embedding_node.get("position", {}) original_position = embedding_node.get("position", {})
@ -421,16 +417,14 @@ class FlowsService:
new_embedding_node["position"] = original_position new_embedding_node["position"] = original_position
# Replace in flow # Replace in flow
self._replace_node_in_flow( self._replace_node_in_flow(flow_data, old_embedding_name, new_embedding_node)
flow_data, old_embedding_id, new_embedding_node
)
components_updated.append( components_updated.append(
f"embedding: {old_embedding_id} -> {new_embedding_id}" f"embedding: {old_embedding_name} -> {new_embedding_id}"
) )
# Replace LLM component (if exists in this flow) # Replace LLM component (if exists in this flow)
if old_llm_id: if old_llm_name:
llm_node = self._find_node_by_id(flow_data, old_llm_id) llm_node, _ = self._find_node_in_flow(flow_data, display_name=old_llm_name)
if llm_node: if llm_node:
# Preserve position # Preserve position
original_position = llm_node.get("position", {}) original_position = llm_node.get("position", {})
@ -440,12 +434,12 @@ class FlowsService:
new_llm_node["position"] = original_position new_llm_node["position"] = original_position
# Replace in flow # Replace in flow
self._replace_node_in_flow(flow_data, old_llm_id, new_llm_node) self._replace_node_in_flow(flow_data, old_llm_name, new_llm_node)
components_updated.append(f"llm: {old_llm_id} -> {new_llm_id}") components_updated.append(f"llm: {old_llm_name} -> {new_llm_id}")
# Replace LLM component (if exists in this flow) # Replace LLM component (if exists in this flow)
if old_llm_text_id: if old_llm_text_name:
llm_text_node = self._find_node_by_id(flow_data, old_llm_text_id) llm_text_node, _ = self._find_node_in_flow(flow_data, display_name=old_llm_text_name)
if llm_text_node: if llm_text_node:
# Preserve position # Preserve position
original_position = llm_text_node.get("position", {}) original_position = llm_text_node.get("position", {})
@ -455,12 +449,18 @@ class FlowsService:
new_llm_text_node["position"] = original_position new_llm_text_node["position"] = original_position
# Replace in flow # Replace in flow
self._replace_node_in_flow( self._replace_node_in_flow(flow_data, old_llm_text_name, new_llm_text_node)
flow_data, old_llm_text_id, new_llm_text_node components_updated.append(f"llm: {old_llm_text_name} -> {new_llm_text_id}")
)
components_updated.append( old_embedding_id = None
f"llm: {old_llm_text_id} -> {new_llm_text_id}" old_llm_id = None
) old_llm_text_id = None
if embedding_node:
old_embedding_id = embedding_node.get("data", {}).get("id")
if old_llm_name and llm_node:
old_llm_id = llm_node.get("data", {}).get("id")
if old_llm_text_name and llm_text_node:
old_llm_text_id = llm_text_node.get("data", {}).get("id")
# Update all edge references using regex replacement # Update all edge references using regex replacement
flow_json_str = json.dumps(flow_data) flow_json_str = json.dumps(flow_data)
@ -478,17 +478,27 @@ class FlowsService:
# Replace LLM ID references (if applicable) # Replace LLM ID references (if applicable)
if old_llm_id: if old_llm_id:
flow_json_str = re.sub(re.escape(old_llm_id), new_llm_id, flow_json_str) flow_json_str = re.sub(
if old_llm_text_id: re.escape(old_llm_id), new_llm_id, flow_json_str
flow_json_str = re.sub( )
re.escape(old_llm_text_id), new_llm_text_id, flow_json_str
)
flow_json_str = re.sub( flow_json_str = re.sub(
re.escape(old_llm_id.split("-")[0]), re.escape(old_llm_id.split("-")[0]),
new_llm_id.split("-")[0], new_llm_id.split("-")[0],
flow_json_str, flow_json_str,
) )
# Replace text LLM ID references (if applicable)
if old_llm_text_id:
flow_json_str = re.sub(
re.escape(old_llm_text_id), new_llm_text_id, flow_json_str
)
flow_json_str = re.sub(
re.escape(old_llm_text_id.split("-")[0]),
new_llm_text_id.split("-")[0],
flow_json_str,
)
# Convert back to JSON # Convert back to JSON
flow_data = json.loads(flow_json_str) flow_data = json.loads(flow_json_str)
@ -510,14 +520,6 @@ class FlowsService:
"flow_id": flow_id, "flow_id": flow_id,
} }
def _find_node_by_id(self, flow_data, node_id):
"""Find a node by ID in the flow data"""
nodes = flow_data.get("data", {}).get("nodes", [])
for node in nodes:
if node.get("id") == node_id:
return node
return None
def _find_node_in_flow(self, flow_data, node_id=None, display_name=None): def _find_node_in_flow(self, flow_data, node_id=None, display_name=None):
""" """
Helper function to find a node in flow data by ID or display name. Helper function to find a node in flow data by ID or display name.
@ -539,14 +541,7 @@ class FlowsService:
return None, None return None, None
async def _update_flow_field( async def _update_flow_field(self, flow_id: str, field_name: str, field_value: str, node_display_name: str = None):
self,
flow_id: str,
field_name: str,
field_value: str,
node_display_name: str = None,
node_id: str = None,
):
""" """
Generic helper function to update any field in any Langflow component. Generic helper function to update any field in any Langflow component.
@ -577,23 +572,18 @@ class FlowsService:
flow_data, display_name=node_display_name flow_data, display_name=node_display_name
) )
if target_node is None and node_id:
target_node, target_node_index = self._find_node_in_flow(
flow_data, node_id=node_id
)
if target_node is None: if target_node is None:
identifier = node_display_name or node_id identifier = node_display_name
raise Exception(f"Component '{identifier}' not found in flow {flow_id}") raise Exception(f"Component '{identifier}' not found in flow {flow_id}")
# Update the field value directly in the existing node # Update the field value directly in the existing node
template = target_node.get("data", {}).get("node", {}).get("template", {}) template = target_node.get("data", {}).get("node", {}).get("template", {})
if template.get(field_name): if template.get(field_name):
flow_data["data"]["nodes"][target_node_index]["data"]["node"]["template"][ flow_data["data"]["nodes"][target_node_index]["data"]["node"]["template"][field_name]["value"] = field_value
field_name if "options" in flow_data["data"]["nodes"][target_node_index]["data"]["node"]["template"][field_name] and field_value not in flow_data["data"]["nodes"][target_node_index]["data"]["node"]["template"][field_name]["options"]:
]["value"] = field_value flow_data["data"]["nodes"][target_node_index]["data"]["node"]["template"][field_name]["options"].append(field_value)
else: else:
identifier = node_display_name or node_id identifier = node_display_name
raise Exception(f"{field_name} field not found in {identifier} component") raise Exception(f"{field_name} field not found in {identifier} component")
# Update the flow via PATCH request # Update the flow via PATCH request
@ -606,41 +596,36 @@ class FlowsService:
f"Failed to update flow: HTTP {patch_response.status_code} - {patch_response.text}" f"Failed to update flow: HTTP {patch_response.status_code} - {patch_response.text}"
) )
async def update_chat_flow_model(self, model_name: str): async def update_chat_flow_model(self, model_name: str, provider: str):
"""Helper function to update the model in the chat flow""" """Helper function to update the model in the chat flow"""
if not LANGFLOW_CHAT_FLOW_ID: if not LANGFLOW_CHAT_FLOW_ID:
raise ValueError("LANGFLOW_CHAT_FLOW_ID is not configured") raise ValueError("LANGFLOW_CHAT_FLOW_ID is not configured")
await self._update_flow_field(
LANGFLOW_CHAT_FLOW_ID,
"model_name",
model_name,
node_display_name="Language Model",
)
async def update_chat_flow_system_prompt(self, system_prompt: str): # Determine target component IDs based on provider
target_llm_id = self._get_provider_component_ids(provider)[1]
await self._update_flow_field(LANGFLOW_CHAT_FLOW_ID, "model_name", model_name,
node_display_name=target_llm_id)
async def update_chat_flow_system_prompt(self, system_prompt: str, provider: str):
"""Helper function to update the system prompt in the chat flow""" """Helper function to update the system prompt in the chat flow"""
if not LANGFLOW_CHAT_FLOW_ID: if not LANGFLOW_CHAT_FLOW_ID:
raise ValueError("LANGFLOW_CHAT_FLOW_ID is not configured") raise ValueError("LANGFLOW_CHAT_FLOW_ID is not configured")
await self._update_flow_field(
LANGFLOW_CHAT_FLOW_ID, # Determine target component IDs based on provider
"system_prompt", target_agent_id = self._get_provider_component_ids(provider)[1]
system_prompt,
node_display_name="Agent", await self._update_flow_field(LANGFLOW_CHAT_FLOW_ID, "system_prompt", system_prompt,
) node_display_name=target_agent_id)
async def update_flow_docling_preset(self, preset: str, preset_config: dict): async def update_flow_docling_preset(self, preset: str, preset_config: dict):
"""Helper function to update docling preset in the ingest flow""" """Helper function to update docling preset in the ingest flow"""
if not LANGFLOW_INGEST_FLOW_ID: if not LANGFLOW_INGEST_FLOW_ID:
raise ValueError("LANGFLOW_INGEST_FLOW_ID is not configured") raise ValueError("LANGFLOW_INGEST_FLOW_ID is not configured")
from config.settings import DOCLING_COMPONENT_ID from config.settings import DOCLING_COMPONENT_DISPLAY_NAME
await self._update_flow_field(LANGFLOW_INGEST_FLOW_ID, "docling_serve_opts", preset_config,
await self._update_flow_field( node_display_name=DOCLING_COMPONENT_DISPLAY_NAME)
LANGFLOW_INGEST_FLOW_ID,
"docling_serve_opts",
preset_config,
node_id=DOCLING_COMPONENT_ID,
)
async def update_ingest_flow_chunk_size(self, chunk_size: int): async def update_ingest_flow_chunk_size(self, chunk_size: int):
"""Helper function to update chunk size in the ingest flow""" """Helper function to update chunk size in the ingest flow"""
@ -664,22 +649,22 @@ class FlowsService:
node_display_name="Split Text", node_display_name="Split Text",
) )
async def update_ingest_flow_embedding_model(self, embedding_model: str): async def update_ingest_flow_embedding_model(self, embedding_model: str, provider: str):
"""Helper function to update embedding model in the ingest flow""" """Helper function to update embedding model in the ingest flow"""
if not LANGFLOW_INGEST_FLOW_ID: if not LANGFLOW_INGEST_FLOW_ID:
raise ValueError("LANGFLOW_INGEST_FLOW_ID is not configured") raise ValueError("LANGFLOW_INGEST_FLOW_ID is not configured")
await self._update_flow_field(
LANGFLOW_INGEST_FLOW_ID,
"model",
embedding_model,
node_display_name="Embedding Model",
)
def _replace_node_in_flow(self, flow_data, old_id, new_node): # Determine target component IDs based on provider
target_embedding_id = self._get_provider_component_ids(provider)[0]
await self._update_flow_field(LANGFLOW_INGEST_FLOW_ID, "model", embedding_model,
node_display_name=target_embedding_id)
def _replace_node_in_flow(self, flow_data, old_display_name, new_node):
"""Replace a node in the flow data""" """Replace a node in the flow data"""
nodes = flow_data.get("data", {}).get("nodes", []) nodes = flow_data.get("data", {}).get("nodes", [])
for i, node in enumerate(nodes): for i, node in enumerate(nodes):
if node.get("id") == old_id: if node.get("data", {}).get("node", {}).get("display_name") == old_display_name:
nodes[i] = new_node nodes[i] = new_node
return True return True
return False return False
@ -734,8 +719,8 @@ class FlowsService:
] ]
# Determine target component IDs based on provider # Determine target component IDs based on provider
target_embedding_id, target_llm_id, target_llm_text_id = ( target_embedding_name, target_llm_name = self._get_provider_component_ids(
self._get_provider_component_ids(provider) provider
) )
results = [] results = []
@ -746,9 +731,8 @@ class FlowsService:
result = await self._update_provider_components( result = await self._update_provider_components(
config, config,
provider, provider,
target_embedding_id, target_embedding_name,
target_llm_id, target_llm_name,
target_llm_text_id,
embedding_model, embedding_model,
llm_model, llm_model,
endpoint, endpoint,
@ -791,24 +775,12 @@ class FlowsService:
def _get_provider_component_ids(self, provider: str): def _get_provider_component_ids(self, provider: str):
"""Get the component IDs for a specific provider""" """Get the component IDs for a specific provider"""
if provider == "watsonx": if provider == "watsonx":
return ( return WATSONX_EMBEDDING_COMPONENT_DISPLAY_NAME, WATSONX_LLM_COMPONENT_DISPLAY_NAME
WATSONX_EMBEDDING_COMPONENT_ID,
WATSONX_LLM_COMPONENT_ID,
WATSONX_LLM_TEXT_COMPONENT_ID,
)
elif provider == "ollama": elif provider == "ollama":
return ( return OLLAMA_EMBEDDING_COMPONENT_DISPLAY_NAME, OLLAMA_LLM_COMPONENT_DISPLAY_NAME
OLLAMA_EMBEDDING_COMPONENT_ID,
OLLAMA_LLM_COMPONENT_ID,
OLLAMA_LLM_TEXT_COMPONENT_ID,
)
elif provider == "openai": elif provider == "openai":
# OpenAI components are the default ones # OpenAI components are the default ones
return ( return OPENAI_EMBEDDING_COMPONENT_DISPLAY_NAME, OPENAI_LLM_COMPONENT_DISPLAY_NAME
OPENAI_EMBEDDING_COMPONENT_ID,
OPENAI_LLM_COMPONENT_ID,
OPENAI_LLM_TEXT_COMPONENT_ID,
)
else: else:
raise ValueError(f"Unsupported provider: {provider}") raise ValueError(f"Unsupported provider: {provider}")
@ -816,9 +788,8 @@ class FlowsService:
self, self,
config, config,
provider: str, provider: str,
target_embedding_id: str, target_embedding_name: str,
target_llm_id: str, target_llm_name: str,
target_llm_text_id: str,
embedding_model: str, embedding_model: str,
llm_model: str, llm_model: str,
endpoint: str = None, endpoint: str = None,
@ -841,7 +812,7 @@ class FlowsService:
# Update embedding component # Update embedding component
if not DISABLE_INGEST_WITH_LANGFLOW: if not DISABLE_INGEST_WITH_LANGFLOW:
embedding_node = self._find_node_by_id(flow_data, target_embedding_id) embedding_node, _ = self._find_node_in_flow(flow_data, display_name=target_embedding_name)
if embedding_node: if embedding_node:
if self._update_component_fields( if self._update_component_fields(
embedding_node, provider, embedding_model, endpoint embedding_node, provider, embedding_model, endpoint
@ -849,22 +820,14 @@ class FlowsService:
updates_made.append(f"embedding model: {embedding_model}") updates_made.append(f"embedding model: {embedding_model}")
# Update LLM component (if exists in this flow) # Update LLM component (if exists in this flow)
if target_llm_id: if target_llm_name:
llm_node = self._find_node_by_id(flow_data, target_llm_id) llm_node, _ = self._find_node_in_flow(flow_data, display_name=target_llm_name)
if llm_node: if llm_node:
if self._update_component_fields( if self._update_component_fields(
llm_node, provider, llm_model, endpoint llm_node, provider, llm_model, endpoint
): ):
updates_made.append(f"llm model: {llm_model}") updates_made.append(f"llm model: {llm_model}")
if target_llm_text_id:
llm_text_node = self._find_node_by_id(flow_data, target_llm_text_id)
if llm_text_node:
if self._update_component_fields(
llm_text_node, provider, llm_model, endpoint
):
updates_made.append(f"llm model: {llm_model}")
# If no updates were made, return skip message # If no updates were made, return skip message
if not updates_made: if not updates_made:
return { return {

38
src/services/langflow_file_service.py
View file

@ -98,7 +98,7 @@ class LangflowFileService:
# Pass metadata via tweaks to OpenSearch component # Pass metadata via tweaks to OpenSearch component
metadata_tweaks = [] metadata_tweaks = []
if owner: if owner or owner is None:
metadata_tweaks.append({"key": "owner", "value": owner}) metadata_tweaks.append({"key": "owner", "value": owner})
if owner_name: if owner_name:
metadata_tweaks.append({"key": "owner_name", "value": owner_name}) metadata_tweaks.append({"key": "owner_name", "value": owner_name})
@ -106,17 +106,18 @@ class LangflowFileService:
metadata_tweaks.append({"key": "owner_email", "value": owner_email}) metadata_tweaks.append({"key": "owner_email", "value": owner_email})
if connector_type: if connector_type:
metadata_tweaks.append({"key": "connector_type", "value": connector_type}) metadata_tweaks.append({"key": "connector_type", "value": connector_type})
logger.info(f"[LF] Metadata tweaks {metadata_tweaks}")
if metadata_tweaks: # if metadata_tweaks:
# Initialize the OpenSearch component tweaks if not already present # # Initialize the OpenSearch component tweaks if not already present
if "OpenSearchHybrid-Ve6bS" not in tweaks: # if "OpenSearchHybrid-Ve6bS" not in tweaks:
tweaks["OpenSearchHybrid-Ve6bS"] = {} # tweaks["OpenSearchHybrid-Ve6bS"] = {}
tweaks["OpenSearchHybrid-Ve6bS"]["docs_metadata"] = metadata_tweaks # tweaks["OpenSearchHybrid-Ve6bS"]["docs_metadata"] = metadata_tweaks
logger.debug( # logger.debug(
"[LF] Added metadata to tweaks", metadata_count=len(metadata_tweaks) # "[LF] Added metadata to tweaks", metadata_count=len(metadata_tweaks)
) # )
if tweaks: if tweaks:
payload["tweaks"] = tweaks payload["tweaks"] = tweaks
logger.debug(f"[LF] Tweaks {tweaks}")
if session_id: if session_id:
payload["session_id"] = session_id payload["session_id"] = session_id
@ -137,9 +138,13 @@ class LangflowFileService:
"X-Langflow-Global-Var-OWNER_EMAIL": str(owner_email), "X-Langflow-Global-Var-OWNER_EMAIL": str(owner_email),
"X-Langflow-Global-Var-CONNECTOR_TYPE": str(connector_type), "X-Langflow-Global-Var-CONNECTOR_TYPE": str(connector_type),
} }
logger.info(f"[LF] Headers {headers}")
logger.info(f"[LF] Payload {payload}")
resp = await clients.langflow_request( resp = await clients.langflow_request(
"POST", f"/api/v1/run/{self.flow_id_ingest}", json=payload, headers=headers "POST",
f"/api/v1/run/{self.flow_id_ingest}",
json=payload,
headers=headers,
) )
logger.debug( logger.debug(
"[LF] Run response", status_code=resp.status_code, reason=resp.reason_phrase "[LF] Run response", status_code=resp.status_code, reason=resp.reason_phrase
@ -160,6 +165,7 @@ class LangflowFileService:
body=resp.text[:1000], body=resp.text[:1000],
error=str(e), error=str(e),
) )
raise raise
return resp_json return resp_json
@ -171,6 +177,10 @@ class LangflowFileService:
settings: Optional[Dict[str, Any]] = None, settings: Optional[Dict[str, Any]] = None,
jwt_token: Optional[str] = None, jwt_token: Optional[str] = None,
delete_after_ingest: bool = True, delete_after_ingest: bool = True,
owner: Optional[str] = None,
owner_name: Optional[str] = None,
owner_email: Optional[str] = None,
connector_type: Optional[str] = None,
) -> Dict[str, Any]: ) -> Dict[str, Any]:
""" """
Combined upload, ingest, and delete operation. Combined upload, ingest, and delete operation.
@ -257,6 +267,10 @@ class LangflowFileService:
session_id=session_id, session_id=session_id,
tweaks=final_tweaks, tweaks=final_tweaks,
jwt_token=jwt_token, jwt_token=jwt_token,
owner=owner,
owner_name=owner_name,
owner_email=owner_email,
connector_type=connector_type,
) )
logger.debug("[LF] Ingestion completed successfully") logger.debug("[LF] Ingestion completed successfully")
except Exception as e: except Exception as e: