RunPod•2mo ago

embeddings endpoints

Hi, I have tried following the sparse documentation, but so far havent been able to get a non-error response or a helpful error message out of the embeddings endpoints. has anyone had any success actually using these, and if so, could you share a setup and exact request format that is known to work?

9 Replies

Keev•5w ago

Following - having the same issue with a ComfyUI workflow. Every time I send a request it just returns: { "delayTime": 8257, "error": "Workflow execution failed: Failed to submit workflow: {"error": "no prompt", "node_errors": []}", "executionTime": 4099, "id": "48d45ad7-66aa-4032-aa24-7f7ee184aed6-e1", "status": "FAILED", "workerId": "djc2dj8t3qkhm9" } Tried numerous different requests and JSON formats

nerdylive•4w ago

Your handler code is wrong, check how to send requests to comfyui again

nerdylive•4w ago

https://github.com/blib-la/runpod-worker-comfy/blob/main/LICENSE you can use this as reference for your comfyui worker @Keev

GitHub

runpod-worker-comfy/LICENSE at main · blib-la/runpod-worker-comfy

ComfyUI as a serverless API on RunPod. Contribute to blib-la/runpod-worker-comfy development by creating an account on GitHub.

nerdylive•4w ago

{
  "input": {
    "model": "YOUR_DEPLOYED_MODEL_NAME",
    "input": "hey, did you know one apple a day that is washed pre-eating is better than unwashed apples?",
  }
}

{
  "input": {
    "model": "YOUR_DEPLOYED_MODEL_NAME",
    "input": "hey, did you know one apple a day that is washed pre-eating is better than unwashed apples?",
  }
}

and you can refer more example here https://github.com/runpod-workers/worker-infinity-embedding?tab=readme-ov-file#openai-compatibility

GitHub

GitHub - runpod-workers/worker-infinity-embedding

Contribute to runpod-workers/worker-infinity-embedding development by creating an account on GitHub.

nerdylive•4w ago

something like that for embeddings @yhlong00000 is it true that embedding workers doesnt need network volumes now ( from the quick deploy menu in serverless )

Keev•2w ago

Got it working, thanks! Having one small issue: I'm not able to generate multiple images from a prompt / request to the endpoint. We have added a variable for the ""batch_size": " value in our workflow, but it only seems to generate one image regardless of the batch_size we give it. This is our Github repo for the runpod worker: https://github.com/sozanski1988/runpod-worker-comfyui/ And the workflow json is here: https://github.com/sozanski1988/runpod-worker-comfyui/blob/main/test_resources/workflows/default_workflow.json Is there a special way that we need to alter the workflow to be able to generate urls for multiple image outputs?

Keev•2w ago

P.S. I've also added an example request and serverless endpoint logs for reference. @nerdylive

logs-runpod-worker-c...

example-request.txt

nerdylive•2w ago

your repo is private, the request seems right (workflow).. you need to chamge your handler code to handle all the files, im guessing it just uploads one file right now find the part where it does that if its true, then change it

Keev•2w ago

thanks you where right, it was the handler. Fixed it

Gaming

Programming

embeddings endpoints

Did you find this page helpful?