{
  "architectures": [
    "MineralNanoForVisionCausalLM"
  ],
  "model_type": "mineral-nano-vision",
  "vocab_size": 32000,
  "hidden_size": 768,
  "intermediate_size": 3072,
  "num_hidden_layers": 12,
  "num_attention_heads": 12,
  "num_key_value_heads": 12,
  "hidden_act": "silu",
  "max_position_embeddings": 2048,
  "initializer_range": 0.02,
  "rms_norm_eps": 1e-6,
  "use_cache": true,
  "pad_token_id": 0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "image_token_id": 32000,
  "tie_word_embeddings": false,
  "rope_theta": 10000.0,
  "rope_scaling": null,
  "attention_bias": false,
  "attention_dropout": 0.0,
  "mlp_bias": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.45.0",
  "vision_config": {
    "hidden_size": 768,
    "image_size": 224,
    "intermediate_size": 3072,
    "num_attention_heads": 12,
    "num_hidden_layers": 12,
    "patch_size": 16,
    "projection_dim": 768,
    "num_channels": 3,
    "layer_norm_eps": 1e-6,
    "dropout": 0.0,
    "attention_dropout": 0.0
  },
  "vision_feature_layer": -2,
  "vision_feature_select_strategy": "default",
  "image_seq_length": 196,
  "num_image_tokens": 196
}