{
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 128000,
  "eos_token_id": [
    128001,
    128008,
    128009
  ],
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 8192,
  "max_position_embeddings": 131072,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 1,
  "num_key_value_heads": 8,
  "pretraining_tp": 1,
  "quantization_config": {
    "bits": 4,
    "example_batch_size": 1,
    "group_size": 256,
    "hadamard_size": 1024,
    "modules_to_not_convert": [
      "lm_head"
    ],
    "p": 2,
    "quant_method": "higgs",
    "tune_metadata": {
      "model.layers.0.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.0.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.0.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.0.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.0.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.0.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.0.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.1.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.1.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.1.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.1.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.1.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.1.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.1.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.10.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.10.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.10.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.10.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.10.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.10.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.10.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.11.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.11.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.11.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.11.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.11.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.11.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.11.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.12.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.12.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.12.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.12.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.12.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.12.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.12.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.13.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.13.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.13.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.13.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.13.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.13.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.13.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.14.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.14.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.14.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.14.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.14.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.14.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.14.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.15.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.15.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.15.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.15.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.15.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.15.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.15.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.2.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.2.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.2.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.2.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.2.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.2.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.2.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.3.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.3.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.3.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.3.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.3.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.3.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.3.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.4.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.4.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.4.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.4.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.4.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.4.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.4.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.5.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.5.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.5.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.5.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.5.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.5.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.5.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.6.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.6.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.6.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.6.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.6.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.6.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.6.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.7.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.7.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.7.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.7.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.7.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.7.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.7.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.8.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.8.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.8.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.8.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.8.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.8.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.8.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.9.mlp.down_proj": {
        "K": 8192,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.9.mlp.gate_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.9.mlp.up_proj": {
        "K": 2048,
        "M": 1,
        "N": 8192,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 136
      },
      "model.layers.9.self_attn.k_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      },
      "model.layers.9.self_attn.o_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.9.self_attn.q_proj": {
        "K": 2048,
        "M": 1,
        "N": 2048,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 88
      },
      "model.layers.9.self_attn.v_proj": {
        "K": 2048,
        "M": 1,
        "N": 512,
        "device": "cuda:0",
        "dtype": "torch.bfloat16",
        "group_size": 256,
        "num_bits": 4,
        "num_sms": 114,
        "template_id": 36
      }
    }
  },
  "rms_norm_eps": 1e-05,
  "rope_scaling": {
    "factor": 32.0,
    "high_freq_factor": 4.0,
    "low_freq_factor": 1.0,
    "original_max_position_embeddings": 8192,
    "rope_type": "llama3"
  },
  "rope_theta": 500000.0,
  "tie_word_embeddings": true,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.52.4",
  "use_cache": true,
  "vocab_size": 128256
}