{
  "A_initializer_range": [
    1,
    16
  ],
  "_name_or_path": "/datadisk1/av11/coding/github/mamba2-torch/models/mamba2-130m-av",
  "architectures": [
    "Mamba2ForCausalLM"
  ],
  "bos_token_id": 0,
  "chunk_size": 256,
  "conv_initializer_range": null,
  "conv_kernel": 4,
  "emb_initializer_range": 0.02,
  "eos_token_id": 0,
  "expand": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 768,
  "intermediate_size": 1536,
  "layer_norm_epsilon": 1e-05,
  "model_type": "mamba2",
  "num_heads": 24,
  "num_hidden_layers": 24,
  "output_last_ssm_states": false,
  "pad_token_id": 0,
  "rescale_prenorm_residual": false,
  "residual_in_fp32": true,
  "state_size": 128,
  "tie_embedding_weights": true,
  "tie_embeddings": true,
  "time_step_floor": 0.0001,
  "time_step_limit": [
    0.0,
    Infinity
  ],
  "time_step_max": 0.1,
  "time_step_min": 0.001,
  "torch_dtype": "float32",
  "transformers_version": "4.42.0.dev0",
  "use_bias": false,
  "use_cache": true,
  "use_conv_bias": true,
  "use_triton_kernels": true,
  "vocab_size": 50288
}