1M-param 4-digit addition model + full results (phase diagram, extrapolation, bake-off)
4a6c57b verified | { | |
| "tag": "sw_1m_4d_50000k", | |
| "size": "1m", | |
| "params": 805632, | |
| "fmt": "cot", | |
| "digits": 4, | |
| "train_tokens": 50000000, | |
| "best_tuning_acc": 1.0, | |
| "best_step": 3660, | |
| "final_bestckpt_acc": 1.0, | |
| "eval_n": 5000, | |
| "wall_min": 2.6284963846206666, | |
| "device": "cuda", | |
| "wrong_examples": [] | |
| } |