config.json

{
  "architectures": [
    "Tipsv2Model"
  ],
  "dtype": "float32",
  "model_type": "tipsv2",
  "temperature_init_value": 0.01,
  "text_config": {
    "attention_dropout": 0.0,
    "bos_token_id": null,
    "eos_token_id": null,
    "hidden_act": "relu",
    "hidden_size": 16,
    "initializer_range": 0.02,
    "intermediate_size": 32,
    "layer_norm_eps": 1e-05,
    "max_position_embeddings": 200,
    "model_type": "tipsv2_text_model",
    "num_attention_heads": 4,
    "num_hidden_layers": 2,
    "pad_token_id": 0,
    "pooling_epsilon": 1e-08,
    "scale_sqrt_depth": true,
    "vocab_size": 32000
  },
  "transformers_version": "5.16.0.dev0",
  "vision_config": {
    "apply_layernorm": true,
    "attention_probs_dropout_prob": 0.0,
    "drop_path_rate": 0.0,
    "hidden_act": "gelu",
    "hidden_dropout_prob": 0.0,
    "hidden_size": 16,
    "image_size": 28,
    "initializer_range": 0.02,
    "layer_norm_eps": 1e-06,
    "layerscale_value": 1.0,
    "mlp_ratio": 2,
    "model_type": "tipsv2_vision_model",
    "num_attention_heads": 4,
    "num_channels": 3,
    "num_hidden_layers": 2,
    "num_register_tokens": 1,
    "out_features": [
      "stage2"
    ],
    "out_indices": [
      2
    ],
    "patch_size": 14,
    "qkv_bias": true,
    "reshape_hidden_states": true,
    "stage_names": [
      "stem",
      "stage1",
      "stage2"
    ],
    "use_swiglu_ffn": false
  }
}