The llm node is not fetching the output from python node .

[dhananjay-nagpal01]

give me insights about why my llm node is not fetching output from my python file. whenever I provide some pdf in input and write prompt in hello.jinja2 file , it doen't answer my query. here is my flow:
$schema: https://azuremlschemas.azureedge.net/promptflow/latest/Flow.schema.json
environment:
python_requirements_txt: requirements.txt
inputs:
text:
type: string
default: C:\Users\dhananjay.nagpaul\Downloads\legal_doc_1.pdf
outputs:
output:
type: string
reference: ${llm.output}
nodes:

• name: python_read
  type: python
  source:
  type: code
  path: python_read.py
  inputs:
  file_path: ${inputs.text}
• name: llm
  type: llm
  source:
  type: code
  path: hello.jinja2
  inputs:
  deployment_name: gpt-35-turbo
  model: gpt-3.5-turbo
  max_tokens: 256
  text: ${python_read}
  connection: to_replace_with_connection_name
  api: chat
• name: hello_prompt
  type: prompt
  source:
  type: code
  path: hello.jinja2
  inputs: {}.

Here is the content of my python node I have created to read the pdf.
from promptflow import tool
import PyPDF2

The inputs section will change based on the arguments of the tool function, after you save the code

Adding type to arguments and return value will help the system show the types properly

Please update the function name/signature per need

@tool
def my_python_tool(pdf_file_path: str) -> str:
pdf_string = ""

# Open pdf file in read binary mode
pdf_file = open(pdf_file_path, 'rb')

# Create PDF reader object
pdf_reader = PyPDF2.PdfReader(pdf_file)

# Read page by page
for page_num in range(len(pdf_reader.pages)):
    page = pdf_reader.pages[page_num]

    # Extract text from page and add to string
    pdf_string += page.extract_text()

# Close the pdf file
pdf_file.close()

return pdf_string

The llm always gives me useless answers that it needs the pdf or it is trying to retrieve the pdf etc. Please help me with this

[0mza987]

In your llm node, please modify text: ${python_read} to text: ${python_read.output}.
Can refer to our example flow dag yamls:
promptflow/examples/flows/standard/web-classification/flow.dag.yaml at main · microsoft/promptflow

[dhananjay-nagpal01]

I have changed that. It is yet now working. LLM is not taking any input from the python node.

[0mza987]

What's the error message?

[0mza987]

Looks like your input file_path in your node python_read does not match the real input name pdf_file_path in your python code. Try to make them the same. And I believe the error message is something like this:

pf.flow.test failed with UserErrorException: TypeError: Execution failure in 'python_read': (TypeError) my_python_tool() got an unexpected keyword argument 'file_path'