{
  "architectures": [
    "HeronForConditionalGeneration"
  ],
  "auto_map": {
    "AutoConfig": "configuration_heron.HeronConfig",
    "AutoModel": "modeling_heron.HeronModel",
    "AutoModelForCausalLM": "modeling_heron.HeronForConditionalGeneration",
    "AutoModelForImageTextToText": "modeling_heron.HeronForConditionalGeneration",
    "AutoModelForVision2Seq": "modeling_heron.HeronForConditionalGeneration"
  },
  "dtype": "bfloat16",
  "image_token_id": 151666,
  "mm_projector_type": "mlp_downsample_2x2_fix",
  "model_type": "heron",
  "text_config": {
    "_name_or_path": "runs/train/NVILA-Lite_0_5b_aws_env2_obelics_ja/sft_0_5b_GPT4_v6_2x2/model/llm",
    "architectures": [
      "Qwen2ForCausalLM"
    ],
    "attention_dropout": 0.0,
    "bos_token_id": 151643,
    "dtype": "bfloat16",
    "eos_token_id": 151645,
    "hidden_act": "silu",
    "hidden_size": 896,
    "initializer_range": 0.02,
    "intermediate_size": 4864,
    "layer_types": [
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention"
    ],
    "max_position_embeddings": 32768,
    "max_window_layers": 21,
    "model_max_length": 4096,
    "model_type": "qwen2",
    "num_attention_heads": 14,
    "num_hidden_layers": 24,
    "num_key_value_heads": 2,
    "rms_norm_eps": 1e-06,
    "rope_scaling": null,
    "rope_theta": 1000000.0,
    "sliding_window": null,
    "tie_word_embeddings": true,
    "use_cache": true,
    "use_sliding_window": false,
    "vocab_size": 151670
  },
  "transformers_version": "4.57.1",
  "vision_config": {
    "_name_or_path": "runs/train/NVILA-Lite_0_5b_aws_env2_obelics_ja/sft_0_5b_GPT4_v6_2x2/model/vision_tower",
    "architectures": [
      "SiglipVisionModel"
    ],
    "attention_dropout": 0.0,
    "dtype": "bfloat16",
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1152,
    "image_size": 448,
    "intermediate_size": 4304,
    "layer_norm_eps": 1e-06,
    "model_type": "siglip_vision_model",
    "num_attention_heads": 16,
    "num_channels": 3,
    "num_hidden_layers": 27,
    "num_image_tokens": 256,
    "patch_size": 14,
    "projection_dim": 2048,
    "projector_hidden_act": "gelu_fast",
    "vision_use_head": false
  },
  "vision_feature_layer": -2
}