Force vLLM to use CUTLASS for FP8 linear (#9)

updated blog post URL

Update Modal account link to use HTTPS

Merge pull request #2 from pipecat-ai/shababo/update-doc-and-deps

test dep compatibility and update docs

temporarily removing pipecat cli because of modal dependency conflict

switch over to point local container to HF models

Sync with gdx-spark-local upstream (50c1771..55af181)

Initial commit