tairov / llama2.mojo

Inference Llama 2 in one file of pure 🔥
https://www.modular.com/blog/community-spotlight-how-i-built-llama2-by-aydyn-tairov
MIT License
2.09k stars 139 forks source link

free all allocated pointers #73

Closed mikowals closed 10 months ago

mikowals commented 10 months ago

I don't notice any impact in the memory use because the Tokenizer is still only deleted just before the program completes. So this is mostly housekeeping in case Tokenizer use changes in the future.

mikowals commented 10 months ago

marked as draft because it looks like running from this branch is producing nonsense output.

mikowals commented 10 months ago

There is no problem with this branch. I had accidentally pulled some broken stuff into my local copy.

tairov commented 10 months ago

Thank you!