{
  "return_dict": true,
  "output_hidden_states": false,
  "dtype": "float16",
  "chunk_size_feed_forward": 0,
  "is_encoder_decoder": false,
  "architectures": [
    "Ovis"
  ],
  "id2label": {
    "0": "LABEL_0",
    "1": "LABEL_1"
  },
  "label2id": {
    "LABEL_0": 0,
    "LABEL_1": 1
  },
  "problem_type": null,
  "_name_or_path": "AIDC-AI/Ovis2-4B",
  "transformers_version": "5.3.0",
  "auto_map": {
    "AutoConfig": "configuration_ovis.OvisConfig",
    "AutoModelForCausalLM": "modeling_ovis.Ovis"
  },
  "model_type": "ovis",
  "llm_config": {
    "vocab_size": 151936,
    "max_position_embeddings": 32768,
    "hidden_size": 2048,
    "intermediate_size": 11008,
    "num_hidden_layers": 36,
    "num_attention_heads": 16,
    "use_sliding_window": false,
    "sliding_window": null,
    "max_window_layers": 70,
    "num_key_value_heads": 2,
    "hidden_act": "silu",
    "initializer_range": 0.02,
    "rms_norm_eps": 1e-06,
    "use_cache": true,
    "attention_dropout": 0.0,
    "layer_types": [
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention",
      "full_attention"
    ],
    "pad_token_id": null,
    "bos_token_id": 151643,
    "eos_token_id": 151645,
    "tie_word_embeddings": true,
    "rope_parameters": {
      "rope_theta": 1000000.0,
      "rope_type": "default"
    },
    "return_dict": true,
    "output_hidden_states": false,
    "dtype": "bfloat16",
    "chunk_size_feed_forward": 0,
    "is_encoder_decoder": false,
    "architectures": [
      "Qwen2ForCausalLM"
    ],
    "id2label": {
      "0": "LABEL_0",
      "1": "LABEL_1"
    },
    "label2id": {
      "LABEL_0": 0,
      "LABEL_1": 1
    },
    "problem_type": null,
    "_name_or_path": "Qwen/Qwen2.5-3B-Instruct",
    "_attn_implementation_autoset": true,
    "add_cross_attention": false,
    "cross_attention_hidden_size": null,
    "decoder_start_token_id": null,
    "finetuning_task": null,
    "is_decoder": false,
    "prefix": null,
    "pruned_heads": {},
    "sep_token_id": null,
    "task_specific_params": null,
    "tf_legacy_loss": false,
    "tie_encoder_decoder": false,
    "tokenizer_class": null,
    "torchscript": false,
    "use_bfloat16": false,
    "model_type": "qwen2",
    "output_attentions": false
  },
  "visual_tokenizer_config": {
    "return_dict": true,
    "output_hidden_states": false,
    "dtype": null,
    "chunk_size_feed_forward": 0,
    "is_encoder_decoder": false,
    "architectures": null,
    "id2label": {
      "0": "LABEL_0",
      "1": "LABEL_1"
    },
    "label2id": {
      "LABEL_0": 0,
      "LABEL_1": 1
    },
    "problem_type": null,
    "_name_or_path": "",
    "_attn_implementation_autoset": true,
    "add_cross_attention": false,
    "backbone_kwargs": {},
    "bos_token_id": null,
    "cross_attention_hidden_size": null,
    "decoder_start_token_id": null,
    "eos_token_id": null,
    "finetuning_task": null,
    "is_decoder": false,
    "pad_token_id": null,
    "prefix": null,
    "pruned_heads": {},
    "sep_token_id": null,
    "task_specific_params": null,
    "tf_legacy_loss": false,
    "tie_encoder_decoder": false,
    "tie_word_embeddings": true,
    "tokenizer_class": null,
    "torchscript": false,
    "use_bfloat16": false,
    "use_indicators": false,
    "vocab_size": 65536,
    "tokenize_function": "softmax",
    "tau": 1.0,
    "depths": null,
    "drop_cls_token": false,
    "backbone_config": {
      "return_dict": true,
      "output_hidden_states": false,
      "dtype": "bfloat16",
      "chunk_size_feed_forward": 0,
      "is_encoder_decoder": false,
      "architectures": [
        "AIMv2Model"
      ],
      "id2label": {
        "0": "LABEL_0",
        "1": "LABEL_1"
      },
      "label2id": {
        "LABEL_0": 0,
        "LABEL_1": 1
      },
      "problem_type": null,
      "_name_or_path": "apple/aimv2-huge-patch14-448",
      "_attn_implementation_autoset": true,
      "add_cross_attention": false,
      "auto_map": {
        "AutoConfig": "configuration_aimv2.AIMv2Config",
        "AutoModel": "modeling_aimv2.AIMv2Model",
        "FlaxAutoModel": "modeling_flax_aimv2.FlaxAIMv2Model"
      },
      "bos_token_id": null,
      "cross_attention_hidden_size": null,
      "decoder_start_token_id": null,
      "eos_token_id": null,
      "finetuning_task": null,
      "is_decoder": false,
      "pad_token_id": null,
      "prefix": null,
      "pruned_heads": {},
      "sep_token_id": null,
      "task_specific_params": null,
      "tf_legacy_loss": false,
      "tie_encoder_decoder": false,
      "tie_word_embeddings": true,
      "tokenizer_class": null,
      "torchscript": false,
      "use_bfloat16": false,
      "hidden_size": 1536,
      "intermediate_size": 4096,
      "num_hidden_layers": 24,
      "num_attention_heads": 12,
      "num_channels": 3,
      "patch_size": 14,
      "image_size": 448,
      "attention_dropout": 0.0,
      "rms_norm_eps": 1e-05,
      "projection_dropout": 0.0,
      "qkv_bias": false,
      "use_bias": false,
      "model_type": "aimv2",
      "output_attentions": false
    },
    "hidden_stride": 2,
    "model_type": "aimv2_visual_tokenizer",
    "output_attentions": false
  },
  "multimodal_max_length": 32768,
  "hidden_size": 2048,
  "conversation_formatter_class": "QwenConversationFormatter",
  "llm_attn_implementation": "flash_attention_2",
  "disable_tie_weight": false,
  "output_attentions": false,
  "quantization_config": {
    "quant_method": "awq",
    "bits": 4,
    "group_size": 128,
    "version": "gemm",
    "zero_point": true,
    "modules_to_not_convert": [
      "visual_tokenizer",
      "vte"
    ]
  }
}