Bulk embedding error

#48
by aikokul - opened

I am getting Failed to Fetch error when I wanted to test batch embedding here: https://jina.ai/embeddings/#faq:~:text=research%20paper%20below.-,Embedding%20API,-Try%20our%20world

Also, i tried to test bulk embedding below. input.csv has only 3 sentences with newlines. But it is raising Error 500: Internal Server Error. I wonder why is this happening and will it be fixed soon?

import requests
import json

File and model details

file_path = "/content/input.csv"
model_name = "jina-embeddings-v3"
email = "[email protected]"

token = userdata.get("JINA_API_KEY")
url = 'https://api.jina.ai/v1/bulk-embeddings'

headers = {"Authorization": f"Bearer {token}"}
data = {'model': model_name, 'email': email}

Send the file in the request

with open(file_path, 'rb') as file:
response = requests.post(url, headers=headers, data=data, files={'file': file})

Handle response

if response.ok:
print("Success:", response.json())
else:
print(f"Error {response.status_code}: {response.text}")

Jina AI org

Hello @aikokul ,

We have identified the problem and we are working on solving it.

However, there is also a bug in your code, it should look something like this:

import requests

# Define the API endpoint
url = 'http://api.jina.ai/v1/bulk-embeddings'

# Define the headers
headers = {
    'accept': 'application/json',
    'Authorization': 'Bearer <jina_api_token>,
}

# Define the files and data for the POST request
files = {
    'file': ('<file_path>', open('<file_path>', 'rb'), 'text/plain'),
}

data = {
    'email': '<your-email>',
    'model': 'jina-embeddings-v3',
}

# Send the POST request
response = requests.post(url, headers=headers, files=files, data=data)

# Output the response
print(response.status_code)
print(response.json())

This is not working yet, will let you know when it does.

Got it, thanks!

Jina AI org

This is now working, Thanks for the patience

Sign up or log in to comment