{ "d_model": 512, "n_heads": 8, "n_kv_heads": 4, "n_layers": 6, "max_seq_len": 512, "device": "cuda", "dataset_files": [ "deepseek_text_20260707_487aef.txt", "deepseek_text_20260707_4f6f18.txt", "deepseek_text_20260707_62ba67.txt", "deepseek_text_20260707_94a4ea.txt", "deepseek_text_20260707_da3020.txt" ] }