fix llama eager runner and add ci

[helunwencser]

Stack from ghstack (oldest at bottom):

• -> fix llama eager runner and add ci #6344
• add ci job for eval_llama with mmlu task #6337
• add CI job for eval_llama with wikitext task #6336

Differential Revision: D64567826

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/6344

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit e06f332 with merge base 8f6c16e ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[helunwencser]

@helunwencser has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

[jackzhxng]

This will be very useful for future llama changes 🙏🏻

[helunwencser]

@helunwencser has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

[helunwencser]

@helunwencser has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

[helunwencser]

@helunwencser has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

[facebook-github-bot]

@helunwencser merged this pull request in 7493aae.