| { | |
| "vocab_size": 125, | |
| "dim": 16, | |
| "n_layers": 7, | |
| "n_heads": 4, | |
| "n_kv_heads": 1, | |
| "intermediate_size": 44, | |
| "max_seq_len": 1024, | |
| "rope_theta": 10000.0, | |
| "rms_norm_eps": 1e-06, | |
| "dropout": 0.1, | |
| "tie_weights": true, | |
| "attention_type": "cubic-strided", | |
| "mod_stride": 16 | |
| } |