docs: fix float16 accumulation data type

[Rohan170603]

What does this PR do?

Related to huggingface/hub-docs#1550

Corrects the accumulation data type for float16 in the quantization concept guide.

The documentation currently lists float16 as both the storage and accumulation data type. This is inconsistent with the surrounding explanation, which describes using a higher-precision accumulation type to avoid overflow and precision loss.

This PR updates:

float16, accumulation data type float16

to:

float16, accumulation data type float32

This is a documentation-only fix; no code changes or tests are required.

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests?

Who can review?

[pcuenca]

looks good to me, thank you!

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.