{
  "model_type": "phenoseq",
  "architecture": "cross-attention-diffusion-denoiser",
  "num_parameters": 168000000
}