config.json

{
  "architectures": [
    "HymbaForCausalLM"
  ],
  "attention_dropout": 0.0,
  "attn_factor": 0.5,
  "attn_hidden_size": 1024,
  "attn_implementation": "flash_attention_2",
  "attn_implementation_new": "flash_attention_2",
  "attn_only_wo_proj": true,
  "auto_map": {
    "AutoConfig": "configuration_hymba.HymbaConfig",
    "AutoModelForCausalLM": "modeling_hymba.HymbaForCausalLM"
  },
  "bos_token_id": 128000,
  "calc_logits_for_entire_prompt": false,
  "conv_dim": {
    "0": 256,
    "1": 256,
    "10": 256,
    "11": 256,
    "12": 256,
    "13": 256,
    "14": 256,
    "15": 256,
    "16": 3200,
    "17": 3200,
    "18": 3200,
    "19": 3200,
    "2": 256,
    "20": 3200,
    "21": 3200,
    "22": 3200,
    "23": 3200,
    "24": 3200,
    "25": 3200,
    "26": 3200,
    "27": 3200,
    "28": 3200,
    "29": 3200,
    "3": 256,
    "30": 3200,
    "31": 3200,
    "4": 256,
    "5": 256,
    "6": 256,
    "7": 256,
    "8": 256,
    "9": 256
  },
  "eos_token_id": 128001,
  "global_attn_idx": [
    0,
    8,
    15
  ],
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 8192,
  "kq_head_dim": -1,
  "kq_norm": "none",
  "kv_reuse_every_i_layer": -1,
  "kv_reuse_group": null,
  "kv_weight_reuse": false,
  "layer_type": [
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h",
    "h"
  ],
  "mamba_conv_bias": true,
  "mamba_d_conv": 4,
  "mamba_d_state": 16,
  "mamba_dt_rank": 100,
  "mamba_expand": 1,
  "mamba_inner_layernorms": false,
  "mamba_proj_bias": false,
  "max_position_embeddings": 131072,
  "memory_tokens_interspersed_every": 0,
  "mlp_hidden_act": "silu",
  "model_type": "hymba",
  "num_attention_heads": 16,
  "num_experts": 1,
  "num_experts_per_tok": 1,
  "num_hidden_layers": 16,
  "num_key_value_heads": 4,
  "num_mamba": 1,
  "num_memory_tokens": 0,
  "orig_max_position_embeddings": 2048,
  "output_router_logits": false,
  "pad_token_id": null,
  "pure_attn": false,
  "repeat_x_before_conv": true,
  "rms_norm_eps": 1e-05,
  "rope": true,
  "rope_scaling": {
    "factor": 32.0,
    "high_freq_factor": 4.0,
    "low_freq_factor": 1.0,
    "original_max_position_embeddings": 8192,
    "rope_type": "llama3"
  },
  "rope_theta": 500000.0,
  "rope_type": "ntk",
  "router_aux_loss_coef": 0.001,
  "seq_length": 8192,
  "sliding_window": 1024,
  "tie_word_embeddings": true,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.53.0",
  "use_cache": false,
  "use_mamba_kernels": true,
  "v_head_dim": 64,
  "vocab_size": 128256
}