| { |
| "problems": 256, |
| "samples": 512, |
| "acc": 0.05859375, |
| "acc_se": 0.011491880758876047, |
| "cut_share": 0.955078125, |
| "mean_tokens": 4019.45703125, |
| "closed_share": 0.0546875, |
| "p_finish": 0.044921875, |
| "p_correct_given_finished": 1.0, |
| "p_correct_given_cut": 0.014314928425357873, |
| "no_answer_finished_share": 0.0, |
| "loop_end_share_of_cut": 0.18200408997955012, |
| "tail_dup8_cut_mean": 0.17756216768916155, |
| "median_tokens_correct": 3307.0, |
| "by_source": { |
| "gsm8k": { |
| "problems": 128, |
| "samples": 256, |
| "acc": 0.07421875, |
| "acc_se": 0.01846889322392738, |
| "cut_share": 0.94140625, |
| "mean_tokens": 3973.06640625 |
| }, |
| "math": { |
| "problems": 128, |
| "samples": 256, |
| "acc": 0.04296875, |
| "acc_se": 0.013615736473364424, |
| "cut_share": 0.96875, |
| "mean_tokens": 4065.84765625 |
| } |
| }, |
| "run": { |
| "name": "dev/base+max4k", |
| "model": "base", |
| "data": "dev256", |
| "n": 2, |
| "shards": 1, |
| "est": 2, |
| "set": [ |
| "max_tokens=4096" |
| ] |
| }, |
| "sampling": { |
| "temperature": 0.6, |
| "top_p": 0.95, |
| "top_k": 20, |
| "presence_penalty": 1.5, |
| "repetition_penalty": 1.0, |
| "max_tokens": 4096, |
| "thinking": true |
| }, |
| "edit_note": null, |
| "tokens_total": 2057962, |
| "gpu_seconds": 192, |
| "tok_per_s_per_gpu": 15439 |
| } |
