config.json

{
  "_name_or_path": "/home/ea/work/my_optimum_intel/optimum-intel/maira2",
  "architectures": [
    "Maira2ForConditionalGeneration"
  ],
  "auto_map": {
    "AutoConfig": "configuration_maira2.Maira2Config",
    "AutoModelForCausalLM": "modeling_maira2.Maira2ForConditionalGeneration",
    "AutoModelForVision2Seq": "modeling_maira2.Maira2ForConditionalGeneration"
  },
  "hidden_size": 16,
  "ignore_index": -100,
  "image_seq_length": 4,
  "image_token_index": 32204,
  "model_type": "maira2",
  "multimodal_projector_bias": true,
  "pad_token_id": 0,
  "projector_hidden_act": "gelu",
  "projector_n_layers": 4,
  "text_config": {
    "_name_or_path": "HuggingFaceM4/tiny-random-LlamaForCausalLM",
    "architectures": [
      "LlamaForCausalLM"
    ],
    "bos_token_id": 0,
    "eos_token_id": 1,
    "head_dim": 4,
    "hidden_size": 16,
    "intermediate_size": 64,
    "model_type": "llama",
    "num_attention_heads": 4,
    "num_hidden_layers": 2,
    "num_key_value_heads": 4,
    "pad_token_id": 2,
    "torch_dtype": "bfloat16",
    "vocab_size": 32207
  },
  "torch_dtype": "float32",
  "transformers_version": "4.48.3",
  "vision_config": {
    "apply_layernorm": true,
    "architectures": [
      "Dinov2Model"
    ],
    "attention_probs_dropout_prob": 0.0,
    "drop_path_rate": 0.0,
    "hidden_act": "gelu",
    "hidden_dropout_prob": 0.0,
    "hidden_size": 16,
    "image_size": 30,
    "layer_norm_eps": 1e-06,
    "layerscale_value": 1.0,
    "mlp_ratio": 4,
    "model_type": "dinov2",
    "num_attention_heads": 4,
    "num_hidden_layers": 4,
    "out_features": [
      "stage4"
    ],
    "out_indices": [
      4
    ],
    "patch_size": 2,
    "qkv_bias": true,
    "reshape_hidden_states": false,
    "stage_names": [
      "stem",
      "stage1",
      "stage2",
      "stage3",
      "stage4"
    ],
    "torch_dtype": "float32",
    "use_swiglu_ffn": false
  },
  "vision_feature_layer": -1,
  "vision_feature_select_strategy": "default"
}