SymbioSLM / server.jl

Commit History

Cache Monarch matrix + causal mask for faster inference
676f15f
Running
verified

LisaMegaWatts commited on

Fix completion_tokens: count tokens not decoded characters
d167fb8
verified

LisaMegaWatts commited on

Upload server.jl with huggingface_hub
62c96ea
verified

LisaMegaWatts commited on