{
  "adaptive": true,
  "architectures": [
    "TransfoXLLMHeadModel"
  ],
  "attn_type": 0,
  "clamp_len": 1000,
  "cutoffs": [
    20000,
    40000,
    200000
  ],
  "d_embed": 1024,
  "d_head": 64,
  "d_inner": 4096,
  "d_model": 1024,
  "div_val": 4,
  "dropatt": 0.2,
  "dropout": 0.2,
  "eos_token_id": 0,
  "init": "normal",
  "init_range": 0.01,
  "init_std": 0.02,
  "layer_norm_epsilon": 1e-05,
  "mem_len": 512,
  "model_type": "transfo-xl",
  "n_head": 16,
  "n_layer": 18,
  "pre_lnorm": false,
  "proj_init_std": 0.01,
  "same_length": true,
  "sample_softmax": -1,
  "tie_projs": [
    false,
    true,
    true,
    true
  ],
  "torch_dtype": "float32",
  "transformers_version": "4.42.3",
  "untie_r": true,
  "vocab_size": 267735
}