{
  "_name_or_path": "fla-hub/gla-7B-mistral-20B",
  "architectures": [
    "GLAForCausalLM"
  ],
  "attn": null,
  "attn_mode": "chunk",
  "bos_token_id": 1,
  "clamp_min": null,
  "conv_size": 4,
  "elementwise_affine": true,
  "eos_token_id": 2,
  "expand_k": 1,
  "expand_v": 1,
  "feature_map": "relu",
  "fuse_cross_entropy": true,
  "fuse_norm": true,
  "hidden_act": "swish",
  "hidden_ratio": 4,
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 14336,
  "max_position_embeddings": 32768,
  "model_type": "gla",
  "norm_eps": 1e-05,
  "num_heads": 32,
  "num_hidden_layers": 32,
  "num_kv_heads": 8,
  "share_conv_kernel": true,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.48.1",
  "use_cache": false,
  "use_gk": true,
  "use_gv": false,
  "use_output_gate": false,
  "use_short_conv": false,
  "vocab_size": 32000
}