config.json

{
  "architectures": [
    "Lfm2MoeCustomForCausalLM"
  ],
  "auto_map": {
  "AutoConfig": "configuration_lfm2_moe_custom.Lfm2MoeCustomConfig",
  "AutoModelForCausalLM": "modeling_lfm2_moe_custom.Lfm2MoeCustomForCausalLM"
  },
  "bos_token_id": 124894,
  "budget": 1,
  "conv_L_cache": 3,
  "conv_bias": false,
  "dtype": "bfloat16",
  "eos_token_id": 124900,
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 7168,
  "layer_types": [
    "conv",
    "conv",
    "full_attention",
    "conv",
    "conv",
    "conv",
    "full_attention",
    "conv",
    "conv",
    "conv",
    "full_attention",
    "conv",
    "conv",
    "conv",
    "full_attention",
    "conv",
    "conv",
    "conv",
    "full_attention",
    "conv",
    "conv",
    "full_attention",
    "conv",
    "conv"
  ],
  "max_position_embeddings": 128000,
  "model_type": "lfm2_moe_custom",
  "moe_intermediate_size": 1792,
  "norm_eps": 1e-05,
  "norm_topk_prob": true,
  "num_attention_heads": 32,
  "num_dense_layers": 2,
  "num_experts": 32,
  "num_experts_per_tok": 4,
  "num_hidden_layers": 24,
  "num_key_value_heads": 8,
  "pad_token_id": 124893,
  "rope_parameters": {
    "rope_theta": 5000000,
    "rope_type": "default"
  },
  "routed_scaling_factor": 1.0,
  "tie_word_embeddings": true,
  "transformers_version": "5.9.0",
  "use_cache": true,
  "use_expert_bias": true,
  "vocab_size": 128000
}