Update sequence_classification.md (#27281)

I'm adding accelerate as one of the libraries to install because otherwise when running the Trainer, the model errorr out with the error. 

ImportError: Using the `Trainer` with `PyTorch` requires `accelerate>=0.20.1`: Please run `pip install transformers[torch]` or `pip install accelerate -U`

Further context: 
1. I've tried this across different environments so I believe that the environment is not the issue. 
2. I had the latest transformers library version running. 
3. Typically even after install accelerate and import it, it wouldn't resolve the issue until I restart the notebook and try again.

docs/source/en/tasks/sequence_classification.md

@@ -44,7 +44,7 @@ The task illustrated in this tutorial is supported by the following model archit
 Before you begin, make sure you have all the necessary libraries installed:
 
 ```bash
-pip install transformers datasets evaluate
+pip install transformers datasets evaluate accelerate
 ```
 
 We encourage you to login to your Hugging Face account so you can upload and share your model with the community. When prompted, enter your token to login: