Mistral-7B_q5, mac M1/M2 binaries

[filopedraz]

• [x] Remove Vicuna & GPT4All services
• [x] Add Mistral service
  • [x] README with docs to download the models & compile the service @casperdcl @filopedraz
  • [ ] Build the docker image for GPU (NVIDIA) @biswaroop1547
  • [x] CI for compiling the code for Mac @casperdcl
  • [x] binary hosting via GH releases
  • [x] merge & manually create a v1 GH release (in future PRs, just force-push v1 tag)
• [x] update registry manifest.json with this spec https://github.com/premAI-io/prem-registry/pull/87:

{
  "weightsDirectoryUrl": "https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF/resolve/main/",
  "weightsFiles" : ["mistral-7b-instruct-v0.1.Q5_0.gguf"],
  "binariesUrl": {
    "aarch64-apple-darwin": "https://github.com/premAI-io/prem-services/releases/download/v1/cht-llama-cpp-mistral-1-aarch64-apple-darwin"
  },
  "serveCommand": "./cht-llama-cpp-mistral-1-aarch64-apple-darwin --model_path=mistral-7b-instruct-v0.1.Q5_0.gguf"
},
{
  "weightsDirectoryUrl": "https://huggingface.co/TheBloke/Mistral-7B-OpenOrca-GGUF/resolve/main/",
  "weightsFiles" : ["mistral-7b-openorca.Q5_K_S.gguf"],
  "binariesUrl": {
    "aarch64-apple-darwin": "https://github.com/premAI-io/prem-services/releases/download/v1/cht-llama-cpp-mistral-1-aarch64-apple-darwin"
  },
  "serveCommand": "./cht-llama-cpp-mistral-1-aarch64-apple-darwin --model_path=mistral-7b-openorca.Q5_K_S.gguf"
}