{
  "model_type": "qovaryx",
  "architectures": [
    "QovaryxForCausalLM"
  ],
  "auto_map": {
    "AutoConfig": "configuration_qovaryx.QovaryxConfig",
    "AutoModelForCausalLM": "modeling_qovaryx.QovaryxForCausalLM"
  },
  "vocab_size": 20242,
  "d_model": 1024,
  "hidden_size": 1024,
  "n_layer": 24,
  "num_hidden_layers": 24,
  "n_head": 16,
  "num_attention_heads": 16,
  "n_kv_head": 4,
  "num_key_value_heads": 4,
  "d_ff": 2816,
  "intermediate_size": 2816,
  "max_seq_len": 2048,
  "max_position_embeddings": 2048,
  "rope_base": 10000.0,
  "rms_eps": 1e-05,
  "dropout": 0.0,
  "decision_head_classes": 4,
  "decision_head_dropout": 0.0,
  "decision_head_enabled": false,
  "mtp_k": 4,
  "mtp_head_kind": "block",
  "init_std": 0.02,
  "tie_word_embeddings": true,
  "ffn_kind": "swiglu",
  "ffn_rank": 128,
  "ffn_experts": 4,
  "ffn_top_k": 1,
  "chart_patch_encoder_enabled": false,
  "chart_image_size": 224,
  "chart_patch_size": 32,
  "chart_channels": 3,
  "chart_embed_dropout": 0.0,
  "torch_dtype": "bfloat16",
  "transformers_version": "5.2.0"
}