wangshiting/linear_llama_math_coding

1{2  "architectures": [3    "LlamaForCausalLM"4  ],5  "attention_bias": false,6  "attention_dropout": 0.0,7  "bos_token_id": 128000,8  "dtype": "float16",9  "eos_token_id": 128009,10  "head_dim": 128,11  "hidden_act": "silu",12  "hidden_size": 3072,13  "initializer_range": 0.02,14  "intermediate_size": 8192,15  "max_position_embeddings": 131072,16  "mlp_bias": false,17  "model_type": "llama",18  "num_attention_heads": 24,19  "num_hidden_layers": 28,20  "num_key_value_heads": 8,21  "pretraining_tp": 1,22  "rms_norm_eps": 1e-05,23  "rope_scaling": {24    "factor": 32.0,25    "high_freq_factor": 4.0,26    "low_freq_factor": 1.0,27    "original_max_position_embeddings": 8192,28    "rope_type": "llama3"29  },30  "rope_theta": 500000.0,31  "tie_word_embeddings": true,32  "transformers_version": "4.57.6",33  "use_cache": false,34  "vocab_size": 12825635}36