config.json

{
  "architectures": [
    "Gemma3ForConditionalGeneration"
  ],
  "boi_token_index": 255999,
  "eoi_token_index": 256000,
  "eos_token_id": [
    1,
    106
  ],
  "image_token_index": 262144,
  "initializer_range": 0.02,
  "mm_tokens_per_image": 256,
  "model_type": "gemma3",
  "quantization_config": {
    "act_bits": 32,
    "amp": true,
    "autoround_version": "0.4.6",
    "batch_size": 1,
    "bits": 4,
    "block_name_to_quantize": "language_model.model.layers",
    "damp_percent": 0.01,
    "data_type": "int",
    "desc_act": false,
    "enable_minmax_tuning": true,
    "enable_norm_bias_tuning": false,
    "enable_quanted_input": true,
    "gradient_accumulate_steps": 8,
    "group_size": 128,
    "iters": 200,
    "low_gpu_mem_usage": false,
    "lr": 0.005,
    "minmax_lr": 0.005,
    "nsamples": 128,
    "quant_method": "auto-round",
    "scale_dtype": "torch.float16",
    "seqlen": 2048,
    "sym": true,
    "true_sequential": false
  },
  "text_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "attn_logit_softcapping": null,
    "cache_implementation": "hybrid",
    "final_logit_softcapping": null,
    "head_dim": 128,
    "hidden_activation": "gelu_pytorch_tanh",
    "hidden_size": 5376,
    "initializer_range": 0.02,
    "intermediate_size": 21504,
    "max_position_embeddings": 131072,
    "model_type": "gemma3_text",
    "num_attention_heads": 32,
    "num_hidden_layers": 62,
    "num_key_value_heads": 16,
    "query_pre_attn_scalar": 168,
    "rms_norm_eps": 1e-06,
    "rope_local_base_freq": 10000.0,
    "rope_scaling": {
      "factor": 8.0,
      "rope_type": "linear"
    },
    "rope_theta": 1000000.0,
    "sliding_window": 1024,
    "sliding_window_pattern": 6,
    "use_cache": true,
    "vocab_size": 262208
  },
  "torch_dtype": "bfloat16",
  "transformers_version": "4.50.0.dev0",
  "vision_config": {
    "attention_dropout": 0.0,
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1152,
    "image_size": 896,
    "intermediate_size": 4304,
    "layer_norm_eps": 1e-06,
    "model_type": "siglip_vision_model",
    "num_attention_heads": 16,
    "num_channels": 3,
    "num_hidden_layers": 27,
    "patch_size": 14,
    "vision_use_head": false
  }
}