config.json

{
  "architectures": [
    "PolyverseForConditionalGeneration"
  ],
  "hidden_size": 768,
  "image_token_index": 128256,
  "model_type": "Polyverse",
  "projection_dim": 2048,
  "text_config": {
    "_name_or_path": "meta-llama/Llama-3.2-1B-Instruct",
    "architectures": [
      "LlamaForCausalLM"
    ],
    "bos_token_id": 128000,
    "eos_token_id": [
      128001,
      128008,
      128009
    ],
    "head_dim": 64,
    "hidden_size": 2048,
    "intermediate_size": 8192,
    "max_position_embeddings": 131072,
    "model_type": "llama",
    "num_hidden_layers": 16,
    "num_image_tokens": 256,
    "num_key_value_heads": 8,
    "pad_token_id": null,
    "rms_norm_eps": 1e-05,
    "rope_scaling": {
      "factor": 32.0,
      "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": true,
    "torch_dtype": "bfloat16",
    "vocab_size": 128258
  },
  "torch_dtype": "float32",
  "transformers_version": "4.44.0",
  "vision_config": {
    "_name_or_path": "google/siglip-base-patch16-256",
    "image_size": 256,
    "model_type": "siglip_vision_model",
    "projection_dim": 2048
  }
}