create_ordering.py not supported with llama 3 loras

[kkailaasa]

Describe the bug

A clear and concise description of what the bug is.

.\mistralrs-server -i lora -o .\llama_lora_ordering.json -m meta-llama/Llama-3.2-1B-Instruct -a jjaegii/Llama-3.2-1B-Instruct-LoRA-ko-kubefix

2024-12-07T00:23:28.343366Z  INFO mistralrs_server: avx: true, neon: false, simd128: false, f16c: true
2024-12-07T00:23:28.343565Z  INFO mistralrs_server: Sampling method: penalties -> temperature -> topk -> topp -> minp -> multinomial                                     
2024-12-07T00:23:28.343692Z  INFO mistralrs_server: Model kind is: lora                                                                                                  
2024-12-07T00:23:28.346944Z  INFO mistralrs_core::pipeline::normal: Loading `tokenizer.json` at `meta-llama/Llama-3.2-1B-Instruct`                                       
2024-12-07T00:23:28.347410Z  INFO mistralrs_core::pipeline::normal: Loading `config.json` at `meta-llama/Llama-3.2-1B-Instruct`                                          
2024-12-07T00:23:30.032688Z  INFO mistralrs_core::pipeline::paths: Found model weight filenames ["model.safetensors"]
Error: Adapter files are empty. Perhaps the ordering file adapters does not match the actual adapters?

llama_lora_ordering.json:

{
    "order": ["jjaegii/Llama-3.2-1B-Instruct-LoRA-ko-kubefix"],
    "layers":
    {
        "model.layers.0.self_attn.q_proj": 0,
        "model.layers.0.self_attn.v_proj": 1,
        "model.layers.1.self_attn.q_proj": 2,
        "model.layers.1.self_attn.v_proj": 3,
        "model.layers.2.self_attn.q_proj": 4,
        "model.layers.2.self_attn.v_proj": 5,
        "model.layers.3.self_attn.q_proj": 6,
        "model.layers.3.self_attn.v_proj": 7,
        "model.layers.4.self_attn.q_proj": 8,
        "model.layers.4.self_attn.v_proj": 9,
        "model.layers.5.self_attn.q_proj": 10,
        "model.layers.5.self_attn.v_proj": 11,
        "model.layers.6.self_attn.q_proj": 12,
        "model.layers.6.self_attn.v_proj": 13,
        "model.layers.7.self_attn.q_proj": 14,
        "model.layers.7.self_attn.v_proj": 15,
        "model.layers.8.self_attn.q_proj": 16,
        "model.layers.8.self_attn.v_proj": 17,
        "model.layers.9.self_attn.q_proj": 18,
        "model.layers.9.self_attn.v_proj": 19,
        "model.layers.10.self_attn.q_proj": 20,
        "model.layers.10.self_attn.v_proj": 21,
        "model.layers.11.self_attn.q_proj": 22,
        "model.layers.11.self_attn.v_proj": 23,
        "model.layers.12.self_attn.q_proj": 24,
        "model.layers.12.self_attn.v_proj": 25,
        "model.layers.13.self_attn.q_proj": 26,
        "model.layers.13.self_attn.v_proj": 27,
        "model.layers.14.self_attn.q_proj": 28,
        "model.layers.14.self_attn.v_proj": 29,
        "model.layers.15.self_attn.q_proj": 30,
        "model.layers.15.self_attn.v_proj": 31
    },
    "base_model_id": "meta-llama/Llama-3.2-1B-Instruct"
}

adapter_config.json of the lora

{
  "alpha_pattern": {},
  "auto_mapping": null,
  "base_model_name_or_path": "meta-llama/Llama-3.2-1B-Instruct",
  "bias": "none",
  "fan_in_fan_out": false,
  "inference_mode": true,
  "init_lora_weights": true,
  "layer_replication": null,
  "layers_pattern": null,
  "layers_to_transform": null,
  "loftq_config": {},
  "lora_alpha": 16,
  "lora_dropout": 0.1,
  "megatron_config": null,
  "megatron_core": "megatron.core",
  "modules_to_save": null,
  "peft_type": "LORA",
  "r": 64,
  "rank_pattern": {},
  "revision": null,
  "target_modules": [
    "q_proj",
    "v_proj"
  ],
  "task_type": "CAUSAL_LM",
  "use_dora": false,
  "use_rslora": false
}

Latest commit or version

Which commit or version you ran with.

v0.3.2

[kkailaasa]

@EricLBuehler if you could take a look and let me know where I am going wrong here... thanks again in advance