CerebrasGPT-256_muP/config.json

{
    "attn_pdrop": 0.0,
    "scale_attn_weights": true,
    "resid_pdrop": 0.0,
    "n_inner": 4352,
    "n_embd": 1088,
    "layer_norm_epsilon": 1e-05,
    "n_positions": 2048,
    "mup_embeddings_scale": 10,
    "activation_function": "gelu",
    "n_head": 17,
    "n_layer": 14,
    "mup_output_alpha": 1.0,
    "mup_width_scale": 0.23529411764705882,
    "position_embedding_type": "learned",
    "mup_scale_qk_dot_by_d": true,
    "tie_word_embeddings": true,
    "vocab_size": 50257,
    "alibi_scaling": null,
    "embd_pdrop": 0.0,
    "model_type": "btlm",
    "use_cache": true,
    "auto_map": {
        "AutoConfig": "cerebras/btlm-3b-8k-base--configuration_btlm.BTLMConfig",
        "AutoModel": "cerebras/btlm-3b-8k-base--modeling_btlm.BTLMModel",
        "AutoModelForCausalLM": "cerebras/btlm-3b-8k-base--modeling_btlm.BTLMLMHeadModel"
    }
}