{
  "_name_or_path": "hf-internal-testing/tiny-xlm-roberta",
  "architectures": [
    "XLMRobertaForCausalLM"
  ],
  "attention_probs_dropout_prob": 0.1,
  "bos_token_id": 0,
  "classifier_dropout": null,
  "d_ff": 256,
  "d_kv": 8,
  "d_model": 64,
  "eos_token_id": 2,
  "gradient_checkpointing": false,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.1,
  "hidden_size": 256,
  "initializer_range": 0.02,
  "intermediate_size": 256,
  "layer_norm_eps": 1e-05,
  "max_position_embeddings": 64,
  "model_type": "xlm-roberta",
  "neuron": {
    "auto_cast": null,
    "auto_cast_type": null,
    "compiler_type": "neuronx-cc",
    "compiler_version": "2.15.128.0+56dc5a86",
    "disable_fallback": false,
    "disable_fast_relayout": false,
    "dynamic_batch_size": false,
    "inline_weights_to_neff": true,
    "input_names": [
      "input_ids",
      "attention_mask"
    ],
    "model_type": "xlm-roberta",
    "optlevel": "2",
    "output_attentions": false,
    "output_hidden_states": false,
    "output_names": [
      "last_hidden_state",
      "pooler_output"
    ],
    "static_batch_size": 1,
    "static_sequence_length": 32
  },
  "num_attention_heads": 2,
  "num_decoder_layers": 2,
  "num_heads": 2,
  "num_hidden_layers": 2,
  "num_layers": 2,
  "output_past": true,
  "pad_token_id": 1,
  "position_embedding_type": "absolute",
  "relative_attention_num_buckets": 32,
  "task": "feature-extraction",
  "torchscript": true,
  "transformers_version": "4.43.2",
  "type_vocab_size": 1,
  "use_cache": true,
  "vocab_size": 5002
}