KoboldAI/Koboldcpp-Tiefighter

Update default.json

Update default.json

Update default.json

Update default.json

Update README.md

Support embedding

Don't hardcode TTS GPU

Update Dockerfile

ggml -> gguf to get rid of a warning

Update default.json

Update default.json

Explicitly add --quiet

Allow secret variables

Fix delimiter

Fix chat completions endpoint

Chat Adapter Support / Debian

Chat / IMG Gen Adapter

Overhaul to use precompiled binaries

Add whisper launch args

attempt to make build work again by building all-major instead of all, and building with more parallel jobs

New Image Gen config style

More efficient ImageGen

Add SD / MMPROJ Args

Make SD / MMPROJ optional

IMG / MMPROJ Support

CUDA 12

API link

Welcome text fix

Space Message

Don't hardcode the title

Make GPU parameters optional and easier to manage

Disable MMQ

Merge branch 'main' of https://huggingface.co/spaces/Henk717/Koboldcpp-Tiefigther

ContextSize as a var

Typo fix

Simplify CMD

MODEL_NAME Fixes

$MODEL_NAME attempt

Better clone-ability

Better clone-ability

Universal Compile

GPU Offlodaing

Expose MODEL Arg

Pull correct distro

GPU Edition

Disable Multiuser

Koboldcpp Tiefigther

initial commit