config.json

{
  "architectures": [
    "XGenMMModelForConditionalGeneration"
  ],
  "auto_map": {
    "AutoConfig": "modeling_xgenmm.XGenMMConfig",
    "AutoModelForVision2Seq": "modeling_xgenmm.XGenMMModelForConditionalGeneration"
  },
  "model_type": "xgenmm",
  "text_config": {
    "attention_dropout": 0.0,
    "embd_pdrop": 0.0,
    "hidden_act": "silu",
    "hidden_size": 3072,
    "initial_tokenizer_len": 32012,
    "initializer_range": 0.02,
    "intermediate_size": 8192,
    "max_position_embeddings": 4096,
    "model_type": "phi3",
    "num_attention_heads": 32,
    "num_hidden_layers": 32,
    "num_key_value_heads": 32,
    "original_max_position_embeddings": 4096,
    "partial_rotary_factor": 1.0,
    "resid_pdrop": 0.0,
    "rms_norm_eps": 1e-05,
    "rope_scaling": null,
    "rope_theta": 10000.0,
    "sliding_window": 2047,
    "torch_dtype": "bfloat16",
    "use_cache": true,
    "vocab_size": 32064
  },
  "torch_dtype": "float32",
  "transformers_version": "4.51.1",
  "vision_encoder_config": {
    "anyres_patch_sampling": true,
    "image_aspect_ratio": "anyres",
    "model_name": "google/siglip-so400m-patch14-384",
    "model_type": "xgenmm_vision_encoder"
  },
  "vision_tokenizer_config": {
    "image_aspect_ratio": "none",
    "lang_embedding_dim": 3072,
    "model_type": "xgenmm_vision_tokenizer",
    "num_vis_tokens": 128,
    "vis_feature_dim": 1152
  }
}