{
  "architectures": [
    "SolarOpen2ForCausalLM"
  ],
  "bos_token_id": 1,
  "eos_token_id": [
    2,
    129
  ],
  "first_k_dense_replace": 0,
  "gqa_interval": 3,
  "gqa_layers": [
    0,
    4,
    8,
    12,
    16,
    20,
    24,
    28,
    32,
    36,
    40,
    44
  ],
  "head_dim": 128,
  "hidden_size": 4096,
  "intermediate_size": 10240,
  "kda_allow_neg_eigval": true,
  "kda_use_full_proj": false,
  "linear_attn_config": {
    "short_conv_kernel_size": 4,
    "head_dim": 128,
    "num_heads": 64,
    "num_kv_heads": null
  },
  "max_position_embeddings": 1048576,
  "model_type": "solar_open2",
  "moe_intermediate_size": 1280,
  "n_routed_experts": 320,
  "n_shared_experts": 1,
  "norm_topk_prob": true,
  "num_attention_heads": 64,
  "num_experts_per_tok": 8,
  "num_hidden_layers": 48,
  "num_key_value_heads": 8,
  "pad_token_id": 2,
  "partial_rotary_factor": 1.0,
  "quantization": {
    "group_size": 64,
    "bits": 4,
    "mode": "affine"
  },
  "quantization_config": {
    "group_size": 64,
    "bits": 4,
    "mode": "affine"
  },
  "rms_norm_eps": 1e-05,
  "rope_theta": 10000,
  "routed_scaling_factor": 1.0,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "5.5.4",
  "use_gqa_gate": true,
  "use_rope": false,
  "vocab_size": 196608,
  "model_file": "solar_open2.py",
  "library_name": "mlx"
}