Make kv cache pos buffer name more specific

[dvorjackz]

Summary

Doing this now just in case later on there appears another module in TorchTune that:

• Uses "cache_pos" as a buffer name
• Appears frequently
• Doesn't require initial state

In this situation we prevent the pte size from blowing up. kv_cache_pos is specific enough that it probably will never appear as the name for another buffer.

Test plan

• Existing unit tests for kv_cache and attention
• E2E
  • python -m examples.models.llama.export_llama --model llama3_2_vision --checkpoint /tmp/Llama-3.2-11B-Vision-Instruct/original/consolidated.pth --params examples/models/llama3_2_vision/text_decoder/params/demo_config.json --metadata '{"append_eos_to_prompt": 0, "get_bos_id":128000, "get_eos_ids":[128009, 128001], "get_n_bos": 0, "get_n_eos": 0}' --output_name="llama3_2_vision.pte" -d fp32 --verbose --max_seq_length 64 -kv
  • python -m examples.models.llama3_2_vision.runner.native --model llama3_2_vision --pte llama3_2_vision.pte --tokenizer /tmp/Llama-3.2-11B-Vision-Instruct/original/tokenizer.model --prompt "Who is the founder of Meta?" --params examples/models/llama3_2_vision/text_decoder/params/demo_config.json --max_len 64 --temperature 0 -kv

Note: I'll make a ci test soon 😅

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/7635

• 📄 Preview Python docs built from this PR

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

⏳ No Failures, 1 Pending

As of commit f369ada with merge base 2a4256b ():
💚 Looks good so far! There are no failures yet. 💚

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