Qualcomm AI Engine Direct - improve llama3.2 3B TPS

[chenweng-quic]

User description

Summary

• Update num_sharding from 4 to 3, TPS change from 26.27 to 28.95.
• Remove dummy Quantize/Dequantize due to tag_ios failure + remove dummy mul 1, TPS change from 28.95 to 29.34.
• Minor fix

before change:

after change:

Minor fix

• Update cli tool option
• Update unit test for spill_fill for newer version of sdk (in newer version of sdk, the spill_fill become 0 and fail the unit test)
• Update unit test root (examples... -> executorch.examples...)

Test plan

python ./examples/qualcomm/oss_scripts/llama/llama.py --artifact llama3_2_3b_instruct --build_folder build-android --soc_model SM8850 --checkpoint ~/.llama/checkpoints/Llama3.2-3B-Instruct/consolidated.00.pth --params ~/.llama/checkpoints/Llama3.2-3B-Instruct/params.json --tokenizer_model ~/.llama/checkpoints/Llama3.2-3B-Instruct/tokenizer.model --prompt "I would like to learn python, could you teach me with a simple example?" --temperature 0 --decoder_model llama3_2-3b_instruct --model_mode kv --max_seq_len 4096 --device ef5e4029 --host localhost --system_prompt "You are a helpful assistant." --max_context_len 4096 --calib_tasks wikitext

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20903

• 📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• No jobs running on PR - "maximum number of queued jobs has been exceeded"

❌ 1 New Failure

As of commit 83581ff with merge base a6d812a ():

NEW FAILURE - The following job has failed:

• pull / test-sqnr-static-llm-qnn-linux (smollm2_135m) / linux-job (gh)
  test_static_llm_model

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

[github-actions]

This PR needs a release notes: label

If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "release notes: none"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.