config.json

{
  "adapter_rank": 128,
  "add_bias_linear": false,
  "architectures": [
    "Zamba2ForCausalLM"
  ],
  "attention_dropout": 0.0,
  "attention_head_dim": 1,
  "attention_hidden_size": 32,
  "bos_token_id": 1,
  "chunk_size": 256,
  "conv_kernel": 3,
  "d_model": 16,
  "d_state": 32,
  "eos_token_id": 2,
  "expand": 2,
  "hidden_act": "gelu",
  "hidden_size": 16,
  "hybrid_layer_ids": [
    2
  ],
  "initializer_range": 0.02,
  "intermediate_size": 64,
  "kv_channels": 0,
  "layers_block_type": [
    "mamba",
    "mamba",
    "hybrid",
    "mamba"
  ],
  "mamba_d_conv": 4,
  "mamba_d_state": 64,
  "mamba_expand": 2,
  "mamba_headdim": 4,
  "mamba_ngroups": 1,
  "max_position_embeddings": 4096,
  "model_type": "zamba2",
  "n_mamba_heads": 8,
  "num_attention_heads": 32,
  "num_hidden_layers": 4,
  "num_key_value_heads": 32,
  "num_logits_to_keep": 1,
  "num_mem_blocks": 1,
  "num_query_groups": 32,
  "pad_token_id": 0,
  "rms_norm_eps": 1e-05,
  "rope_theta": 10000,
  "time_step_floor": 0.0001,
  "time_step_limit": null,
  "time_step_max": 0.1,
  "time_step_min": 0.001,
  "torch_dtype": "float32",
  "transformers_version": "4.55.4",
  "use_cache": true,
  "use_conv_bias": true,
  "use_long_context": false,
  "use_mem_eff_path": false,
  "use_mem_rope": false,
  "use_shared_attention_adapter": false,
  "vocab_size": 50280
}