{
  "architectures": [
    "EmbeddingGemma2Model"
  ],
  "audio_config": {
    "_name_or_path": "",
    "architectures": null,
    "attention_chunk_size": 12,
    "attention_context_left": 13,
    "attention_context_right": 0,
    "attention_invalid_logits_value": -1000000000.0,
    "attention_logit_cap": 50.0,
    "chunk_size_feed_forward": 0,
    "conv_kernel_size": 5,
    "dtype": "float32",
    "gradient_clipping": 10000000000.0,
    "hidden_act": "silu",
    "hidden_size": 32,
    "id2label": {
      "0": "LABEL_0",
      "1": "LABEL_1"
    },
    "initializer_range": 0.02,
    "is_encoder_decoder": false,
    "label2id": {
      "LABEL_0": 0,
      "LABEL_1": 1
    },
    "model_type": "gemma4_audio",
    "num_attention_heads": 2,
    "num_hidden_layers": 2,
    "output_attentions": false,
    "output_hidden_states": false,
    "output_proj_dims": 48,
    "problem_type": null,
    "residual_weight": 0.5,
    "return_dict": true,
    "rms_norm_eps": 1e-06,
    "subsampling_conv_channels": [
      128,
      4
    ],
    "use_clipped_linears": true
  },
  "audio_token_id": 258881,
  "boa_token_id": 256000,
  "boi_token_id": 255999,
  "dtype": "float32",
  "eoa_token_index": 258883,
  "eoi_token_id": 258882,
  "image_token_id": 258880,
  "initializer_range": 0.02,
  "model_type": "embedding_gemma2",
  "text_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "bos_token_id": 2,
    "dtype": "float32",
    "embedding_dim": 24,
    "eos_token_id": 1,
    "head_dim": 8,
    "hidden_activation": "gelu_pytorch_tanh",
    "hidden_size": 32,
    "hidden_size_per_layer_input": 8,
    "initializer_range": 0.02,
    "intermediate_size": 64,
    "layer_types": [
      "sliding_attention",
      "sliding_attention",
      "full_attention",
      "sliding_attention",
      "sliding_attention",
      "full_attention"
    ],
    "max_position_embeddings": 262144,
    "model_type": "embedding_gemma2_text",
    "num_attention_heads": 4,
    "num_hidden_layers": 6,
    "num_key_value_heads": 2,
    "pad_token_id": 0,
    "per_layer_config": {
      "2": {
        "head_dim": 16,
        "num_key_value_heads": 1
      },
      "5": {
        "head_dim": 16,
        "num_key_value_heads": 1
      }
    },
    "rms_norm_eps": 1e-06,
    "rope_parameters": {
      "full_attention": {
        "rope_theta": 1000000.0,
        "rope_type": "default"
      },
      "sliding_attention": {
        "rope_theta": 10000.0,
        "rope_type": "default"
      }
    },
    "sliding_window": 4,
    "vocab_size": 262144
  },
  "transformers_version": "5.18.0.dev0",
  "video_token_id": 258884,
  "vision_config": {
    "_name_or_path": "",
    "architectures": null,
    "attention_bias": false,
    "attention_dropout": 0.0,
    "chunk_size_feed_forward": 0,
    "default_output_length": 280,
    "dtype": "float32",
    "global_head_dim": 16,
    "head_dim": 16,
    "hidden_activation": "gelu_pytorch_tanh",
    "hidden_size": 32,
    "id2label": {
      "0": "LABEL_0",
      "1": "LABEL_1"
    },
    "initializer_range": 0.02,
    "intermediate_size": 64,
    "is_encoder_decoder": false,
    "label2id": {
      "LABEL_0": 0,
      "LABEL_1": 1
    },
    "max_position_embeddings": 131072,
    "model_type": "gemma4_vision",
    "num_attention_heads": 2,
    "num_hidden_layers": 2,
    "num_key_value_heads": 2,
    "output_attentions": false,
    "output_hidden_states": false,
    "patch_size": 16,
    "pooling_kernel_size": 3,
    "position_embedding_size": 10240,
    "problem_type": null,
    "return_dict": true,
    "rms_norm_eps": 1e-06,
    "rope_parameters": {
      "rope_theta": 100.0,
      "rope_type": "axial"
    },
    "standardize": false,
    "use_clipped_linears": false
  },
  "vision_soft_tokens_per_image": 280,
  "transformers.js_config": {
    "use_external_data_format": {
      "audio_encoder.onnx": 1,
      "model.onnx": 1,
      "model_fp16.onnx": 1,
      "vision_encoder.onnx": 1,
      "vision_encoder_fp16.onnx": 1
    }
  }
}