Datasets:
v2 and v3 previews: tasks, references and replay-verified GPT-6.1 Sol pilot episodes; training summary (part 3)
ff57e68 verified | [ | |
| { | |
| "checkpoint_step": 25, | |
| "n_tasks": 50, | |
| "n": 50, | |
| "reward_mean": 0.5652, | |
| "reward_std": 0.3114, | |
| "valid_rate": 1.0, | |
| "tool_call_rate": 1.0, | |
| "feasible_rate": 0.74, | |
| "rate_ok": 0.74, | |
| "rate_all_idle": 0.0, | |
| "rate_overfill": 0.1, | |
| "rate_sell_before_charge": 0.08, | |
| "rate_rate": 0.06, | |
| "rate_cycle_cap": 0.02, | |
| "rate_end_short": 0.0, | |
| "rate_malformed": 0.0, | |
| "rate_truncated": 0.0, | |
| "completion_len_mean": 244.8, | |
| "ci95": [ | |
| 0.4756, | |
| 0.6481 | |
| ], | |
| "by_start": { | |
| "empty": { | |
| "n": 18, | |
| "reward_mean": 0.5789, | |
| "ci95": [ | |
| 0.4045, | |
| 0.7446 | |
| ] | |
| }, | |
| "full": { | |
| "n": 19, | |
| "reward_mean": 0.502, | |
| "ci95": [ | |
| 0.3664, | |
| 0.6349 | |
| ] | |
| }, | |
| "half": { | |
| "n": 13, | |
| "reward_mean": 0.6384, | |
| "ci95": [ | |
| 0.5243, | |
| 0.7409 | |
| ] | |
| } | |
| }, | |
| "by_curriculum": { | |
| "C0": { | |
| "n": 17, | |
| "reward_mean": 0.6043, | |
| "ci95": [ | |
| 0.4493, | |
| 0.7525 | |
| ] | |
| }, | |
| "C1": { | |
| "n": 10, | |
| "reward_mean": 0.4284, | |
| "ci95": [ | |
| 0.2534, | |
| 0.6012 | |
| ] | |
| }, | |
| "C2": { | |
| "n": 23, | |
| "reward_mean": 0.5957, | |
| "ci95": [ | |
| 0.4716, | |
| 0.7171 | |
| ] | |
| } | |
| } | |
| }, | |
| { | |
| "checkpoint_step": 50, | |
| "n_tasks": 50, | |
| "n": 50, | |
| "reward_mean": 0.5655, | |
| "reward_std": 0.3131, | |
| "valid_rate": 1.0, | |
| "tool_call_rate": 1.0, | |
| "feasible_rate": 0.76, | |
| "rate_ok": 0.76, | |
| "rate_all_idle": 0.0, | |
| "rate_overfill": 0.12, | |
| "rate_sell_before_charge": 0.06, | |
| "rate_rate": 0.06, | |
| "rate_cycle_cap": 0.0, | |
| "rate_end_short": 0.0, | |
| "rate_malformed": 0.0, | |
| "rate_truncated": 0.0, | |
| "completion_len_mean": 237.7, | |
| "ci95": [ | |
| 0.476, | |
| 0.6488 | |
| ], | |
| "by_start": { | |
| "empty": { | |
| "n": 18, | |
| "reward_mean": 0.5966, | |
| "ci95": [ | |
| 0.4385, | |
| 0.7481 | |
| ] | |
| }, | |
| "full": { | |
| "n": 19, | |
| "reward_mean": 0.474, | |
| "ci95": [ | |
| 0.3423, | |
| 0.6043 | |
| ] | |
| }, | |
| "half": { | |
| "n": 13, | |
| "reward_mean": 0.6561, | |
| "ci95": [ | |
| 0.4972, | |
| 0.7971 | |
| ] | |
| } | |
| }, | |
| "by_curriculum": { | |
| "C0": { | |
| "n": 17, | |
| "reward_mean": 0.656, | |
| "ci95": [ | |
| 0.5139, | |
| 0.7875 | |
| ] | |
| }, | |
| "C1": { | |
| "n": 10, | |
| "reward_mean": 0.3622, | |
| "ci95": [ | |
| 0.1966, | |
| 0.5362 | |
| ] | |
| }, | |
| "C2": { | |
| "n": 23, | |
| "reward_mean": 0.5869, | |
| "ci95": [ | |
| 0.4595, | |
| 0.7091 | |
| ] | |
| } | |
| } | |
| } | |
| ] |