config.json

{
  "architectures": [
    "LeanMixtralForCausalLM"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoModelForCausalLM": "modeling_lean_mixtral.LeanMixtralForCausalLM"
  },
  "bos_token_id": 1,
  "dtype": "bfloat16",
  "eos_token_id": 2,
  "head_dim": null,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 14336,
  "leanmix_compressed_kv_layers": [
    3,
    8,
    13,
    14,
    16,
    22,
    23,
    24,
    25,
    26,
    27,
    28
  ],
  "leanmix_kv_decoder_depth": {
    "13": 2,
    "14": 2,
    "16": 2,
    "22": 2,
    "23": 2,
    "24": 2,
    "25": 2,
    "26": 2,
    "27": 2,
    "28": 2,
    "3": 2,
    "8": 2
  },
  "leanmix_kv_decoder_hidden_dim": {
    "13": 128,
    "14": 128,
    "16": 128,
    "22": 128,
    "23": 128,
    "24": 128,
    "25": 128,
    "26": 128,
    "27": 128,
    "28": 128,
    "3": 128,
    "8": 128
  },
  "leanmix_kv_fixed_alpha": {
    "13": 1.0,
    "14": 1.0,
    "16": 1.0,
    "22": 1.0,
    "23": 1.0,
    "24": 1.0,
    "25": 1.0,
    "26": 1.0,
    "27": 1.0,
    "28": 1.0,
    "3": 1.0,
    "8": 1.0
  },
  "leanmix_kv_input_dims": {
    "13": 1024,
    "14": 1024,
    "16": 1024,
    "22": 1024,
    "23": 1024,
    "24": 1024,
    "25": 1024,
    "26": 1024,
    "27": 1024,
    "28": 1024,
    "3": 1024,
    "8": 1024
  },
  "leanmix_kv_key_dims": {
    "13": 4,
    "14": 4,
    "16": 4,
    "22": 4,
    "23": 4,
    "24": 4,
    "25": 4,
    "26": 4,
    "27": 4,
    "28": 4,
    "3": 4,
    "8": 4
  },
  "leanmix_kv_residual_decoder": {
    "13": false,
    "14": false,
    "16": false,
    "22": false,
    "23": false,
    "24": false,
    "25": false,
    "26": false,
    "27": false,
    "28": false,
    "3": false,
    "8": false
  },
  "leanmix_kv_value_dims": {
    "13": 4,
    "14": 4,
    "16": 4,
    "22": 4,
    "23": 4,
    "24": 4,
    "25": 4,
    "26": 4,
    "27": 4,
    "28": 4,
    "3": 4,
    "8": 4
  },
  "max_position_embeddings": 32768,
  "model_type": "mixtral",
  "num_attention_heads": 32,
  "num_experts_per_tok": 2,
  "num_hidden_layers": 32,
  "num_key_value_heads": 8,
  "num_local_experts": 8,
  "output_router_logits": false,
  "rms_norm_eps": 1e-05,
  "rope_theta": 1000000.0,
  "router_aux_loss_coef": 0.02,
  "router_jitter_noise": 0.0,
  "sliding_window": null,
  "tie_word_embeddings": false,
  "transformers_version": "4.57.6",
  "use_cache": true,
  "vocab_size": 32000
}