config.json

{
  "architectures": [
    "CacaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "attention_pattern": "all_global",
  "attention_sink_size": 4,
  "attention_sink_window": 1024,
  "attn_logit_softcapping": null,
  "audio_config": {},
  "auto_map": {
    "AutoConfig": "caca_transformers.CacaConfig",
    "AutoModel": "caca_transformers.CacaModel",
    "AutoModelForCausalLM": "caca_transformers.CacaForCausalLM"
  },
  "bos_token_id": 1,
  "chat_template": "{% for message in messages %}{% if message['role'] == 'system' %}System: {{ message['content'] }}\n{% elif message['role'] == 'user' %}User: {{ message['content'] }}\n{% elif message['role'] == 'assistant' %}Assistant: {{ message['content'] }}\n{% endif %}{% endfor %}{% if add_generation_prompt %}Assistant:{% endif %}",
  "cross_attention_frequency": 4,
  "dtype": "float32",
  "eos_token_id": 2,
  "expert_capacity_factor": 1.0,
  "expert_choice_k": 0.125,
  "final_logit_softcapping": null,
  "global_attention_every_n_layers": 2,
  "head_dim": 64,
  "hidden_dropout": 0.1,
  "hidden_size": 1024,
  "initializer_range": 0.02,
  "intermediate_size": 4096,
  "layer_scale_init": 1e-05,
  "longformer_attention_window": 512,
  "max_position_embeddings": 8192,
  "merge_threshold": 0.5,
  "mlp_bias": false,
  "mod_capacity_factor": 0.5,
  "mod_route_method": "learned",
  "model_type": "caca",
  "moe_layer_frequency": 2,
  "num_attention_heads": 16,
  "num_expert_groups": 1,
  "num_experts": 8,
  "num_experts_per_tok": 2,
  "num_hidden_layers": 24,
  "num_key_value_heads": 4,
  "pipeline_parallel_size": 1,
  "pretraining_tp": 1,
  "projector_hidden_size": 1024,
  "qk_norm_eps": 1e-06,
  "residual_dropout": 0.1,
  "rms_norm_eps": 1e-06,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "router_aux_loss_coef": 0.01,
  "router_z_loss_coef": 0.001,
  "sliding_window": null,
  "stochastic_depth_prob": 0.1,
  "tensor_parallel_size": 1,
  "tie_word_embeddings": false,
  "transformers_version": "4.57.3",
  "use_alibi": false,
  "use_attention_sink": false,
  "use_cache": true,
  "use_cross_attention": false,
  "use_expert_choice": false,
  "use_flash_attn": true,
  "use_grouped_moe": false,
  "use_grouped_query_attention": false,
  "use_layer_scale": false,
  "use_longformer_attention": false,
  "use_mixture_of_depths": false,
  "use_moe": false,
  "use_multi_query_attention": false,
  "use_multimodal": false,
  "use_qk_norm": true,
  "use_rotary_embeddings": true,
  "use_soft_merging": false,
  "use_stochastic_depth": false,
  "vision_config": {},
  "vocab_size": 32000
}