migrated to groq -- suuuuuper fast!

Update requirements.txt

fix for the fix

some modifications, this is hacky. I'm trying to get some tangible result first and then will polish everything for production

a fix which may work

reqs

test-llama3, not sure how it'll go on langserve

[add-model] llama-8b-instruct