| { |
| "depth": 12, |
| "aspect_ratio": 64, |
| "head_dim": 128, |
| "max_seq_len": 2048, |
| "vocab_size": 16384, |
| "params_total": 185598434, |
| "params_scaling": 97518000, |
| "flops_per_token": 811600560, |
| "unique_train_tokens": 183061543, |
| "target_tokens_ratio": 1023939000, |
| "target_tokens_epochs": 732246172, |
| "horizon": "branch x1.5 of the stable run", |
| "num_iterations": 3605, |
| "total_batch_size": 262144, |
| "device_batch_size": 4, |
| "world_size": 1, |
| "device_batch_size_source": "ladder.json device_batch_size_by_vram", |
| "grad_accum_steps": 32, |
| "train_tokens": 945029120, |
| "epochs": 5.16, |
| "tokens_per_param": 9.69, |
| "train_flops": 818260411353661440, |
| "window_pattern": "L", |
| "attention": "sdpa", |
| "fp8": false, |
| "eval_tokens": 3719168, |
| "eval_every": 90, |
| "save_every": 180, |
| "keep_checkpoints": true, |
| "base_overrides": {}, |
| "warmdown_ratio": 0.2, |
| "stream": true, |
| "decay_at": 2884, |
| "plain_at": 3245, |
| "loader_tails": true, |
| "regularization": { |
| "wd_schedule": "floor", |
| "wd_floor": 0.5, |
| "table_wd_mult": 1.0, |
| "dropout": 0.0, |
| "dropout_from": 0.5 |
| }, |
| "dropout_from_step": null, |
| "wd_horizon": 3846, |
| "save_steps": [], |
| "seed": null, |
| "arm": "R1", |
| "branches": { |
| "1": { |
| "decay_at": 1923, |
| "decay_steps": 481, |
| "num_iterations": 2404 |
| }, |
| "1.5": { |
| "decay_at": 2884, |
| "decay_steps": 721, |
| "num_iterations": 3605 |
| }, |
| "2": { |
| "decay_at": 3846, |
| "decay_steps": 962, |
| "num_iterations": 4808 |
| } |
| }, |
| "tag": "queen-v4-r1-x1.5", |
| "branch": { |
| "of": "queen-v4-r1-stable", |
| "at": 2884, |
| "decay_steps": 721 |
| } |
| } |
