{ "d_k": 64, "d_model": 1024, "dropout_prob": 0.1, "from_zero": true, "gqa_div": 1, "l": 128, "max_len": 2048, "n_heads": 16, "n_layers": 24, "norm_eps": 1e-05, "share_embeds": true, "vocab_size": 32768 }