Getting error back from API Key. I will attach my code below. I'm also not seeing any console log messages, and my failed attempts are charging me a few cents. Need advice.

[furerjos]

Here is my code. I am unable to see any console log messages, yet a file is being produced and charging me a few cents. Trying to iterate a prompt over a file with 300 rows, testing it on just 2 rows for now. Working on an independent project, changed the names of the fields for the sake of security for my idea. I'm confused why I don't see any console logs, and why I keep getting errors back for the two fields I want Perplexity to produce: Date and Summary. Any advice? Thank you so much for your time, this is my first ever community post. Tried following documentation, even using the prompt in the Getting Started doesn't work. I pip installed OpenAI by the way.

from openai import OpenAI
import pandas as pd
import time
import sys

API setup

YOUR_API_KEY = "nuhUhUh!" # Replace with your API key
client = OpenAI(api_key=YOUR_API_KEY, base_url="https://api.perplexity.ai")

File paths

input_file = r"fileWithDataForPrompts"
output_file = r"OutputFile"

Load the Excel file

try:
data = pd.read_excel(input_file)
print(f"Loaded {len(data)} rows from the input file.")
except Exception as e:
print(f"Failed to load input file: {e}")
sys.exit(1)

Validate headers

required_headers = ["Enitity", "EntityId", "Name", "Amount", "Date"]
if not all(header in data.columns for header in required_headers):
raise ValueError(f"The input file must contain the headers: {required_headers}")

Results list

results = []

Testing with the first 2 rows

data_sample = data.head(2)

Process rows

for index, row in data_sample.iterrows():
try:
print(f"Processing row {index + 1}/{len(data_sample)}...", flush=True)

    # Construct messages
    messages = [
        {"role": "system", "content": "You are a helpful assistant for summarizing details."},
        {
            "role": "user",
            "content": (
                f"Please summarize the following contract details:\n"
                f"Entity: {row['Entity']}, "
                f"EntityId: {row['EntityId']}, "
                f"Name: {row['Name']}, "
                f"Amount: {row['Amount']}, "
                f"Date: {row['Date']}. "
                "Include any details such as scope, completion date, and key features."
            ),
        },
    ]

    # API request with streaming
    response_content = ""
    response_stream = client.chat.completions.create(
        model="llama-3.1-sonar-large-128k-online",
        messages=messages,
        stream=True,
    )
    for chunk in response_stream:
        print(f"Chunk received: {chunk}", flush=True)
        response_content += chunk["choices"][0]["delta"].get("content", "")

    # Extract summary and completion date
    summary = response_content.split("\n")[0]
    completion_date = "Not available"
    for line in response_content.split("\n"):
        if "completion" in line.lower() or "expiration" in line.lower():
            completion_date = line.strip()
            break

    # Append results
    results.append({
        "Entity": row["Entity"],
        "EntityId": row["EntityId"],
        "Name": row["Name"],
        "Amount": row["Amount"],
        "Date": row["Date"],
        "Summary": summary,
        "Completion/Expiration Date": completion_date,
    })
    print(f"Row {index + 1} processed: Summary and date extracted.", flush=True)

except Exception as e:
    print(f"Error for row {index + 1}: {e}", flush=True)
    results.append({
        "Agency": row["Agency"],
        "Contract Number": row["Contract Number"],
        "Contractor": row["Contractor"],
        "Award Amount": row["Award Amount"],
        "Award Date": row["Award Date"],
        "Summary": "Error",
        "Completion/Expiration Date": "Error",
    })

time.sleep(2)  # Prevent API rate limits

Save results

try:
pd.DataFrame(results).to_excel(output_file, index=False)
print(f"Results saved to {output_file}", flush=True)
except Exception as e:
print(f"Failed to save results: {e}", flush=True)

[aarashy]

Hey, we never charge you if the API call fails, but it's not obvious to me if theres a bug in your python post-processing