Qualcomm AI Engine Direct - improve llama3.2 3B TPS (#20903)

### **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:
<img width="991" height="242" alt="image"
src="https://github.com/user-attachments/assets/0d04eeb9-0934-405e-a21c-e2bf31a77211"
/>


after change:
<img width="1045" height="242" alt="image"
src="https://github.com/user-attachments/assets/b7992e63-1677-4e8a-a974-3702e4d12d4c"
/>

#### 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
```