# URGENT HELP on Endpoint invokation

**URL:** <https://discuss.huggingface.co/t/urgent-help-on-endpoint-invokation/26009>\
**Category:** 🤗Transformers\
**Created:** [November 13, 2022, 5:31pm UTC](https://discuss.huggingface.co/t/urgent-help-on-endpoint-invokation/26009 "2022-11-13T17:31:23Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![thanksfinance](https://avatars.discourse-cdn.com/v4/letter/t/ea666f/32.png) [@thanksfinance](https://discuss.huggingface.co/u/thanksfinance)\
**Post date:** [November 13, 2022, 5:31pm UTC](https://discuss.huggingface.co/t/urgent-help-on-endpoint-invokation/26009/1 "2022-11-13T17:31:23Z")

</div>

I’m using a SageMaker endpoint with _cardiffnlp/twitter-roberta-base-sentiment-latest._

When I Invoke the endpoint my input is:

```auto
        input = {
            "inputs": texts, #It's a List of texts
            "parameters": {
                "truncation": True 
                }
            }

```

It was running smoothly for days… and then I received:

The expanded size of the tensor (679) must match the existing size (514) at non-singleton dimension 1. Target sizes: [1, 679]. Tensor sizes: [1, 514]

How can I fix?

---

<div class="post-metadata">

**Author:** ![thanksfinance](https://avatars.discourse-cdn.com/v4/letter/t/ea666f/32.png) [@thanksfinance](https://discuss.huggingface.co/u/thanksfinance)\
**Post date:** [November 13, 2022, 5:46pm UTC](https://discuss.huggingface.co/t/urgent-help-on-endpoint-invokation/26009/2 "2022-11-13T17:46:57Z")

</div>

max\_length: 510 Works!