config.json

{
  "architectures": [
    "RicottaForCausalLM"
  ],
  "auto_map": {
    "AutoConfig": "configuration_ricotta.RicottaConfig",
    "AutoModelForCausalLM": "modeling_ricotta.RicottaForCausalLM"
  },
  "dtype": "float32",
  "max_seq_len": 4096,
  "mlp_hidden": 768,
  "model_type": "gouda-ricotta",
  "n_embd": 256,
  "n_head": 4,
  "n_kv_head": 2,
  "n_layer": 20,
  "name": "Ricotta-2.0",
  "qk_norm": true,
  "rms_eps": 1e-06,
  "rope_base": 10000.0,
  "tie_word_embeddings": true,
  "transformers_version": "5.17.0",
  "vocab_size": 32768,
  "z_loss": 0.0
}