Fix EXAONE-4.0 dummy id

[lkm2835]

What does this PR do?

Fixes #39129 (comment)
LGAI-EXAONE/EXAONE-4.0-Instruct -> LGAI-EXAONE/EXAONE-4.0-32B

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a Github issue or the forum? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

@ydshieh

[ydshieh]

run-slow: exaone4

[github-actions]

This comment contains run-slow, running the specified jobs:

models: ['models/exaone4']
quantizations: [] ...

[github-actions]

[For maintainers] Suggested jobs to run (before merge)

run-slow: exaone4

[lkm2835]

Hi @ydshieh
Apologies, I’ve updated the commit and pushed it again.

[ydshieh]

Hi @lkm2835 I haven't check the report of the previous commit. What is wrong with the 1.2B model?

(if we can make it work with smaller model, it is even better)

[lkm2835]

I agree that testing with the smaller model is lighter and faster.
However, the 32B model includes features such as No Positional Embedding and local–global hybrid attention that are not present in the 1.2B architecture.
For better test coverage, it would be better to run tests on the 32B model.

[ydshieh]

run-slow: exaone4

[github-actions]

This comment contains run-slow, running the specified jobs:

models: ['models/exaone4']
quantizations: [] ...

[ydshieh]

The process is being killed

tests/models/exaone4/test_modeling_exaone4.py::Exaone4IntegrationTest::test_model_generation
-------------------------------- live log call ---------------------------------
WARNING accelerate.big_modeling:big_modeling.py:442 Some parameters are on the meta device because they were offloaded to the cpu.
WARNING transformers.generation.configuration_utils:logging.py:328 The following generation flags are not valid and may be ignored: ['temperature']. Set TRANSFORMERS_VERBOSITY=info for more details.
FAILED [ 97%]
Killed

https://github.com/huggingface/transformers/actions/runs/17949641363/job/51045205804

I will check tomorrow and proposal some fixes.

(But if you want to explore a solution before that, feel free to try)

We are on a runner of A10 (24 G GPU ram and 60 G CPM ram)