{
  "architectures": [
    "FineViTModel"
  ],
  "auto_map": {
    "AutoConfig": "config.FineViTConfig",
    "AutoModel": "finevit.FineViTModel"
  },
  "backbone_config": {
    "_name_or_path": "facebook/dinov2-with-registers-base",
    "add_cross_attention": false,
    "apply_layernorm": true,
    "architectures": [
      "Dinov2WithRegistersModel"
    ],
    "attention_probs_dropout_prob": 0.0,
    "bos_token_id": null,
    "chunk_size_feed_forward": 0,
    "cross_attention_hidden_size": null,
    "decoder_start_token_id": null,
    "drop_path_rate": 0.0,
    "dtype": "float32",
    "eos_token_id": null,
    "finetuning_task": null,
    "hidden_act": "gelu",
    "hidden_dropout_prob": 0.0,
    "hidden_size": 768,
    "id2label": {
      "0": "LABEL_0",
      "1": "LABEL_1"
    },
    "image_size": 518,
    "initializer_range": 0.02,
    "interpolate_antialias": true,
    "interpolate_offset": 0.0,
    "is_decoder": false,
    "is_encoder_decoder": false,
    "label2id": {
      "LABEL_0": 0,
      "LABEL_1": 1
    },
    "layer_norm_eps": 1e-06,
    "layerscale_value": 1.0,
    "mlp_ratio": 4,
    "model_type": "dinov2_with_registers",
    "num_attention_heads": 12,
    "num_channels": 3,
    "num_hidden_layers": 12,
    "num_register_tokens": 4,
    "out_features": [
      "stage12"
    ],
    "out_indices": [
      12
    ],
    "output_attentions": false,
    "output_hidden_states": false,
    "pad_token_id": null,
    "patch_size": 14,
    "prefix": null,
    "problem_type": null,
    "pruned_heads": {},
    "qkv_bias": true,
    "reshape_hidden_states": true,
    "return_dict": true,
    "sep_token_id": null,
    "stage_names": [
      "stem",
      "stage1",
      "stage2",
      "stage3",
      "stage4",
      "stage5",
      "stage6",
      "stage7",
      "stage8",
      "stage9",
      "stage10",
      "stage11",
      "stage12"
    ],
    "task_specific_params": null,
    "tf_legacy_loss": false,
    "tie_encoder_decoder": false,
    "tie_word_embeddings": true,
    "tokenizer_class": null,
    "torchscript": false,
    "transformers_version": "5.8.0",
    "use_bfloat16": false,
    "use_swiglu_ffn": false
  },
  "backbone_model_name": "facebook/dinov2-with-registers-base",
  "checkpoint_epoch": 2,
  "checkpoint_step": 50000,
  "dropout": 0.0,
  "dtype": "bfloat16",
  "image_size": 224,
  "init_backbone_from_pretrained": false,
  "initializer_range": 0.02,
  "latent_encoder_head_dim": 64,
  "latent_encoder_mlp_hidden_dim": 3072,
  "latent_encoder_num_attention_heads": 12,
  "latent_encoder_num_layers": 12,
  "model_type": "finevit",
  "num_latents": 64,
  "patch_decoder_head_dim": 64,
  "patch_decoder_mlp_hidden_dim": 3072,
  "patch_decoder_num_attention_heads": 12,
  "patch_decoder_num_layers": 8,
  "teacher_encoder_type": "dinov2_with_registers",
  "train_encoder": true,
  "transformers_version": "5.8.0"
}