config.json

{
  "activation_fn_name": "swish",
  "architectures": [
    "TpttModel"
  ],
  "auto_map": {
    "AutoConfig": "configuration_tptt.TpttConfig",
    "AutoModelForCausalLM": "modeling_tptt.TpttModel"
  },
  "base_model_name": "apple/OpenELM-1_1B",
  "ffn_dim_divisor": 256,
  "ffn_multipliers": [
    0.5,
    0.63,
    0.76,
    0.89,
    1.02,
    1.15,
    1.28,
    1.41,
    1.54,
    1.67,
    1.8,
    1.93,
    2.06,
    2.19,
    2.31,
    2.44,
    2.57,
    2.7,
    2.83,
    2.96,
    3.09,
    3.22,
    3.35,
    3.48,
    3.61,
    3.74,
    3.87,
    4.0
  ],
  "ffn_with_glu": true,
  "head_dim": 64,
  "initializer_range": 0.02,
  "lora_config": {
    "alpha_pattern": {},
    "auto_mapping": null,
    "base_model_name_or_path": null,
    "bias": "none",
    "eva_config": null,
    "exclude_modules": null,
    "fan_in_fan_out": false,
    "inference_mode": false,
    "init_lora_weights": true,
    "layer_replication": null,
    "layers_pattern": null,
    "layers_to_transform": null,
    "loftq_config": {},
    "lora_alpha": 16,
    "lora_bias": false,
    "lora_dropout": 0.05,
    "megatron_config": null,
    "megatron_core": "megatron.core",
    "modules_to_save": null,
    "peft_type": "LORA",
    "r": 8,
    "rank_pattern": {},
    "revision": null,
    "target_modules": [
      "out_proj",
      "qkv_proj"
    ],
    "task_type": "CAUSAL_LM",
    "use_dora": false,
    "use_rslora": false
  },
  "mag_weight": 0.5,
  "max_chunk_size": 64,
  "max_context_length": 2048,
  "max_self_attn_length": 8192,
  "model_dim": 2048,
  "model_type": "tptt",
  "normalization_layer_name": "rms_norm",
  "normalize_qk_projections": true,
  "num_gqa_groups": 4,
  "num_kv_heads": [
    4,
    4,
    4,
    5,
    5,
    5,
    5,
    5,
    5,
    5,
    6,
    6,
    6,
    6,
    6,
    6,
    6,
    6,
    7,
    7,
    7,
    7,
    7,
    7,
    8,
    8,
    8,
    8
  ],
  "num_query_heads": [
    16,
    16,
    16,
    20,
    20,
    20,
    20,
    20,
    20,
    20,
    24,
    24,
    24,
    24,
    24,
    24,
    24,
    24,
    28,
    28,
    28,
    28,
    28,
    28,
    32,
    32,
    32,
    32
  ],
  "num_transformer_layers": 28,
  "operator_mode": "delta_rule",
  "qkv_multipliers": [
    0.5,
    1.0
  ],
  "rope_freq_constant": 10000,
  "rope_max_length": 4096,
  "share_input_output_layers": true,
  "target_modules_names": [
    "attn",
    "self_attn",
    "attention"
  ],
  "torch_dtype": "bfloat16",
  "transformers_version": "4.51.3",
  "use_cache": true,
  "vocab_size": 32000
}