{
  "_name_or_path": "mistralai/Mistral-7B-v0.1",
  "architectures": [
    "SparseMistralforCausalLM"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "sparsification_sftt.SparseMistralConfig",
    "AutoModelForCausalLM": "sparsification_sftt.SparseMistralforCausalLM"
  },
  "bos_token_id": 1,
  "cut_pre_attn": false,
  "cut_pre_mlp": false,
  "eos_token_id": 2,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 14336,
  "max_position_embeddings": 32768,
  "model_type": "sparse_mistral",
  "num_attention_heads": 32,
  "num_hidden_layers": 32,
  "num_key_value_heads": 8,
  "rms_norm_eps": 1e-05,
  "rope_theta": 10000.0,
  "sliding_window": 4096,
  "thresholds": [
    0.0247037410736084,
    0.03390511870384216,
    0.04450671002268791,
    0.05650851130485535,
    0.07611145079135895,
    0.09411414712667465,
    0.10531583428382874,
    0.11991801857948303,
    0.1259189248085022,
    0.12691907584667206,
    0.1307196468114853,
    0.13372008502483368,
    0.1375206559896469,
    0.1405211091041565,
    0.15412315726280212,
    0.16792522370815277,
    0.18552786111831665,
    0.19032858312129974,
    0.20693106949329376,
    0.2187328338623047,
    0.22653400897979736,
    0.2371356040239334,
    0.23433518409729004,
    0.2371356040239334,
    0.24213634431362152,
    0.24213634431362152,
    0.240136057138443,
    0.24213634431362152,
    0.24313649535179138,
    0.240136057138443,
    0.2371356040239334,
    0.22753416001796722
  ],
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.36.2",
  "use_cache": false,
  "use_relu": false,
  "use_resilu": false,
  "use_sparse_model": true,
  "use_sparse_predictor": false,
  "use_sparse_regularization": false,
  "vocab_size": 32000
}