config.json

{
  "architectures": [
    "LlavaLlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 128000,
  "eos_token_id": [
    128001,
    128008,
    128009
  ],
  "hidden_act": "silu",
  "hidden_size": 4096,
  "image_aspect_ratio": "anyres",
  "image_grid_pinpoints": [
    [
      1344,
      448
    ],
    [
      2240,
      448
    ],
    [
      896,
      896
    ],
    [
      448,
      2688
    ],
    [
      448,
      1344
    ],
    [
      448,
      4032
    ],
    [
      3136,
      448
    ],
    [
      1792,
      896
    ],
    [
      1344,
      1344
    ],
    [
      4032,
      448
    ],
    [
      896,
      1792
    ],
    [
      448,
      896
    ],
    [
      896,
      448
    ],
    [
      448,
      2240
    ],
    [
      2688,
      448
    ],
    [
      448,
      3584
    ],
    [
      1344,
      896
    ],
    [
      1792,
      448
    ],
    [
      3584,
      448
    ],
    [
      448,
      1792
    ],
    [
      896,
      1344
    ],
    [
      448,
      3136
    ]
  ],
  "initializer_range": 0.02,
  "intermediate_size": 14336,
  "max_position_embeddings": 131072,
  "mlp_bias": false,
  "mm_hidden_size": 3200,
  "mm_patch_merge_type": "spatial_unpad",
  "mm_projector": {
    "add_image_newline_embed": false,
    "add_pos_embed": false,
    "conv_block_depth": 2,
    "downsample_rate": 4,
    "downsample_size": [
      16,
      16
    ],
    "enable_gradient_checkpointing": false,
    "mlp_depth": 2,
    "num_input_token": 1024,
    "trainable": true,
    "type": "CAbstractorMMProjector"
  },
  "mm_projector_lr": 2e-06,
  "mm_use_im_patch_token": false,
  "mm_use_im_start_end": false,
  "mm_vision_tower": {
    "mm_vision_select_feature": "patch",
    "mm_vision_select_layer": -1,
    "pretrained_model_name_or_path": "/cache/data/huggingface_models/InternViT-6B-448px-V1-5",
    "trainable": false,
    "type": "InternVisionTower",
    "unfreeze_mm_vision_tower": true
  },
  "model_type": "llava_llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 32,
  "num_key_value_heads": 8,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": {
    "factor": 8.0,
    "high_freq_factor": 4.0,
    "low_freq_factor": 1.0,
    "original_max_position_embeddings": 8192,
    "rope_type": "llama3"
  },
  "rope_theta": 500000.0,
  "tie_word_embeddings": false,
  "tokenizer_model_max_length": 4096,
  "tokenizer_padding_side": "right",
  "torch_dtype": "bfloat16",
  "transformers_version": "4.44.2",
  "use_cache": false,
  "use_mm_proj": true,
  "vocab_size": 132353
}