{
  "architectures": [
    "PixtralLarge4VisionModel"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "modeling_pixtral_large4.PixtralLarge4VisionConfig",
    "AutoModel": "modeling_pixtral_large4.PixtralLarge4VisionModel"
  },
  "dtype": "bfloat16",
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "image_size": 1540,
  "initializer_range": 0.02,
  "intermediate_size": 6144,
  "model_type": "pixtral_large4_vision",
  "num_attention_heads": 16,
  "num_channels": 3,
  "num_hidden_layers": 30,
  "patch_size": 14,
  "rope_parameters": {
    "rope_theta": 10000.0,
    "rope_type": "axial"
  },
  "transformers_version": "5.17.0"
}