config.json

{
  "architectures": [
    "Lfm2ForCausalLM"
  ],
  "block_auto_adjust_ff_dim": true,
  "block_dim": 640,
  "block_ffn_dim_multiplier": 1.0,
  "block_mlp_init_scale": 1.0,
  "block_multiple_of": 256,
  "block_norm_eps": 1e-05,
  "block_out_init_scale": 1.0,
  "block_use_swiglu": true,
  "block_use_xavier_init": true,
  "bos_token_id": 1,
  "conv_L_cache": 3,
  "conv_bias": false,
  "conv_dim": 640,
  "conv_dim_out": 640,
  "conv_use_xavier_init": true,
  "dtype": "float32",
  "eos_token_id": 2,
  "full_attn_idxs": [
    2,
    5,
    8,
    10
  ],
  "hidden_size": 640,
  "initializer_range": 0.02,
  "intermediate_size": 2560,
  "layer_types": [
    "conv",
    "conv",
    "full_attention",
    "conv",
    "conv",
    "full_attention",
    "conv",
    "conv",
    "full_attention",
    "conv",
    "full_attention",
    "conv"
  ],
  "max_position_embeddings": 32768,
  "model_type": "lfm2",
  "norm_eps": 1e-05,
  "num_attention_heads": 10,
  "num_heads": 10,
  "num_hidden_layers": 12,
  "num_key_value_heads": 5,
  "pad_token_id": 3,
  "rope_parameters": {
    "rope_theta": 500000.0,
    "rope_type": "default"
  },
  "tie_word_embeddings": false,
  "transformers_version": "5.8.1",
  "use_cache": false,
  "use_pos_enc": true,
  "vocab_size": 32768
}