config.json

{
  "audio_hidden_size": 384,
  "audio_vocab_size": 1024,
  "bos_token_id": 2,
  "causal_reasoning_layers": 2,
  "chunk_size": 128,
  "context_size": 1024,
  "d_model": 768,
  "eos_token_id": 3,
  "image_vocab_size": 8192,
  "intermediate_size": 3072,
  "max_audio_tokens": 500,
  "max_image_tokens": 256,
  "model_type": "sykoomni",
  "n_heads": 6,
  "n_layers": 24,
  "num_global_memory_tokens": 32,
  "num_memory_tokens": 16,
  "pad_token_id": 0,
  "text_vocab_size": 32000,
  "transformers_version": "5.2.0",
  "vision_hidden_size": 768,
  "vocab_size": 41226
}