Information Extractor node, 2 outputs

Questions
1

ISSUE

I’m using the Ollama Model node linked to the Information Extraction node
When I execute, I get two outputs from the ollama node.

They start with two different inputs based on “format_instructions” prompt.
For this test case, I don’t have enabled any prompt in the Information Extraction node, and the Output Format option is also not enabled in the ollama node.

The input of the first output reads:

System: You are an expert extraction algorithm.
Only extract relevant information from the text.
If you do not know the value of an attribute asked to extract, you may omit the attribute's value.
You must format your output as a JSON value that adheres to a given "JSON Schema" instance.

"JSON Schema" is a declarative language that allows you to annotate and validate JSON documents.

For example, the example "JSON Schema" instance {{"properties": {{"foo": {{"description": "a list of test words", "type": "array", "items": {{"type": "string"}}}}}}, "required": ["foo"]}}}}
would match an object with one required property, "foo". The "type" property specifies "foo" must be an "array", and the "description" property semantically describes it as "a list of test words". The items within "foo" must be strings.
Thus, the object {{"foo": ["bar", "baz"]}} is a well-formatted instance of this example "JSON Schema". The object {{"properties": {{"foo": ["bar", "baz"]}}}} is not well-formatted.

Your output will be parsed and type-checked according to the provided schema instance, so make sure all fields in your output match the schema exactly and there are no trailing commas!

Here is the JSON Schema instance your output must adhere to. Include the enclosing markdown codeblock:
```json
{"type":"object","properties":{"comercio":{"type":"string"},"fecha":{"type":"string","format":"date"},"items":...


The input of the second output of the ollama node reads the same, except for the beggining:

Instructions:
--------------
You must format your output as a JSON value that adheres to a given "JSON Schema" instance.

"JSON Schema" is a declarative language that allows you to annotate and validate JSON documents.
...

I’m not sure if I’m missing something here, the process indeed is slow since the LLM is churning tokens.
Any help is welcome.

By the way, for the ollama base url I had to change localhost with http://127.0.0.1:11434 in order for it to work.

n8n 1.75.2 local host on Win11.
ollama cli 0.5.7 / lmstudio 0.3.8-2
llama3.2:3b-instruct-q8_0
2

It looks like your topic is missing some important information. Could you provide the following if applicable.

  • n8n version:
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app):
  • Operating system:
3

Now also testing with LMStudio, the log says it received two messages:

2025-01-27 09:17:47  [INFO] 
[LM STUDIO SERVER] Running chat completion on conversation with 2 messages.
2025-01-27 09:17:47  [INFO] 
[LM STUDIO SERVER] Accumulating tokens ... (stream = false)
2025-01-27 09:18:47  [INFO] 
[LM STUDIO SERVER] Client disconnected. Stopping generation... (If the model is busy processing the prompt, it will finish first.)
2025-01-27 09:18:47  [INFO] 
[LM STUDIO SERVER] [llama-3.2-3b-instruct] Generated prediction:  {

Then after a few minutes, the workflow times out with an error:

{
  "errorMessage": "Request timed out.",
  "errorDetails": {
    "rawErrorMessage": [
      "Request timed out."
    ]
  },
  "n8nDetails": {
    "nodeName": "OpenAI Chat Model",
    "nodeType": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
    "nodeVersion": 1,
    "time": "27/1/2025, 9:19:48",
    "n8nVersion": "1.75.2 (Self Hosted)",
    "binaryDataMode": "default",
    "stackTrace": [
      "NodeApiError: Request timed out.",
      "    at Object.onFailedAttempt (C:\\AI\\n8n\\node_modules\\@n8n\\n8n-nodes-langchain\\nodes\\llms\\n8nLlmFailedAttemptHandler.ts:40:20)",
      "    at RetryOperation._fn (C:\\AI\\n8n\\node_modules\\p-retry\\index.js:67:20)",
      "    at processTicksAndRejections (node:internal/process/task_queues:105:5)"
    ]
  }
}
4

Then by chance it works again, with either ollama or lmstudio.
In this case Information Extractor only sends one prompt, a joint prompt from mine and the format_instructions prompt.

It’s very erratic and the Model node doesn’t “wake” the model timely, it has a hard time connecting through the API end point.

If you guys can check please. I was thinking on self-hosting n8n with ollama, or maybe use openrouter, but this is not reliable enough for production.

Thank you.

5

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.