{
  "architectures": [
    "Glm4vVisionModel"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "depth": 2,
  "dtype": "float32",
  "hidden_act": "silu",
  "hidden_size": 48,
  "image_size": 336,
  "in_channels": 3,
  "initializer_range": 0.02,
  "intermediate_size": 22,
  "model_type": "glm4v_vision",
  "num_heads": 12,
  "out_hidden_size": 16,
  "patch_size": 14,
  "rms_norm_eps": 1e-05,
  "spatial_merge_size": 1,
  "temporal_patch_size": 2,
  "transformers_version": "5.16.0.dev0"
}