name: test_model best_checkpoint_path: ./outputs/checkpoints/test_model/best.pt latest_checkpoint_path: ./outputs/checkpoints/test_model/latest.pt tokenizer: bpe_8k d_model: 512 num_heads: 4 num_layers: 4 d_ff: 1024 dropout_rate: 0.1 vocab_size: 8192 max_seq_len: 128 pad_id: 0 bos_id: 1 eos_id: 2 layer_norm_style: post