Error: Enable to run Lora - Adapter files are empty

[kkailaasa]

Describe the bug

when trying to run mistral server with a lora repeatedly getting Error: Adapter files are empty. Perhaps the ordering file adapters does not match the actual adapters?

have tried this on different systems and also changing to different file paths and checking - in any permutation combination still getting the same issue

Folder Dir:
Directory structure for: \test_lora

├── config.toml
├── mistralrs-server.exe
├── ordering.json
└── [sample_data_adapter]
    ├── README.md
    ├── adapter_config.json
    ├── adapter_model.safetensors
    ├── special_tokens_map.json
    ├── tokenizer.json
    └── tokenizer_config.json

ordering.json file:

{
    "order": ["sample_data"],
    "layers": {
        "model.layers.0.self_attn.q_proj": 0,
        "model.layers.0.self_attn.k_proj": 1,
        "model.layers.0.self_attn.v_proj": 2,
        "model.layers.0.self_attn.o_proj": 3,
        "model.layers.0.mlp.gate_proj": 4,
        "model.layers.0.mlp.up_proj": 5,
        "model.layers.0.mlp.down_proj": 6
    },
    "base_model_id": "meta-llama/Llama-3.2-1B-Instruct",
    "preload_adapters": [{
        "name": "spiritual_love",
        "adapter_model_id": "sample_data_adapter"
    }]
}

cmds tried:
./mistralrs-server -i lora --model-id meta-llama/Llama-3.2-1B-Instruct --adapters-model-id ./sample_data_adapter --order ./ordering.json

mistralrs-server -i toml -f lora_config.toml (when using toml, only base model ran without the lora)

both for linux and windows, mistralrs-server was built from source

Latest commit or version

running version 0.3.2/git pull as of 22_11_24

[kkailaasa]

Hello again, just reverting back on this - if anyone could check and help me out - Please and thank you..