| { | |
| "dataset_sha256": "a701c0b511adae786e7895e954bece8eae427a6054ad64cc9c35865419054529", | |
| "model": { | |
| "vocab_size": 274, | |
| "d_model": 512, | |
| "nhead": 8, | |
| "num_layers": 6, | |
| "dim_feedforward": 3072, | |
| "max_seq_len": 128, | |
| "dropout": 0.1 | |
| }, | |
| "kind": "two_tower", | |
| "batch": 4096, | |
| "epochs": 100, | |
| "warmup": 2000, | |
| "seed": 42, | |
| "world": 4, | |
| "compile": true, | |
| "zero_key_bias": true | |
| } |