inference server and endpoints for elle-elle-aime

[monperrus]

A Triton inference server might be useful for the open-source models

https://github.com/triton-inference-server

[monperrus]

inference server by huggingface https://github.com/huggingface/text-generation-inference

we may get an instance soon with StarCoder

[monperrus]

@GGmorello has set up RepairLLama over HuggingFace Spaces thanks to our zero-gpus account

[monperrus]

@FredBonux is able to use Mixtral and LLama over groq for free, see https://www.groq.com