{
  "architectures": [
    "PerceiverCausalLanguageModel"
  ],
  "is_decoder": true,
  "model_config": {
    "abs_pos_emb": false,
    "activation_checkpointing": false,
    "activation_offloading": false,
    "cross_attention_dropout": 0.0,
    "cross_attention_widening_factor": 4,
    "init_scale": 0.02,
    "max_heads_parallel": 2,
    "max_latents": 512,
    "max_seq_len": 1024,
    "num_channels": 1280,
    "num_heads": 10,
    "num_self_attention_layers": 20,
    "output_bias": false,
    "output_norm": true,
    "post_attention_dropout": 0.0,
    "self_attention_widening_factor": 4,
    "vocab_size": 32000
  },
  "model_type": "perceiver-ar-causal-language-model",
  "tokenizer_class": "XLNetTokenizerFast",
  "torch_dtype": "float32",
  "transformers_version": "4.28.0"
}