两臂 n=256 同 seed 背靠背(配方=轮次42 K8+cert_batch=1+inc_cap=1):到达 0.5234→0.5039 (-1.95pp,逐题配对丢6赢1净-5,McNemar 精确 p=0.125 与零不可分),时限内 +1.17pp、 刹停 -9%、行程比 -0.076、执行碰撞仍为 0(判据未放松)。丢的 6 题 5 题在 A 类(开阔区) =买余量卖进度的签名。 仪表与离线数交叉验证:改选率 5.58% 对上轮次44 独立测得的"执行点间隙落在 0.2–0.5mm 占 5.6%",并列率 79.0% 对上"满前缀 79.4%" ⇒ 分辨率上界被宽相位截顶 gcap=max(0.5,margin+0.5) 锁死,5.6% 是这个家族在 cap=0.5 下的天花板,不是实现没调好。 结构性发现:#24 作用在错的人群——tie 只在 maxlen>0 触发,而失败主人群是全拒(maxlen=0, 未到达题 22% 的规划);全拒那条已由 backoff(轮次43)判掉。前置条件记为 #28(gcap 与 gap_margin 解耦)。 实现账:新增 tie 与触发计数器(生产端 per 数组+dump 键、消费端汇总 tie_* 同批改完, tie=0 时 22/22 老键逐位复现);修掉 tie 分支局部名 nmax 顶掉主循环总步数的遮蔽 bug, 守卫单测做过变异检查;计数器正对照可手算(tie_opp=tie_changed=n_plans)。47 项测试绿。
112 lines
2.7 KiB
JSON
112 lines
2.7 KiB
JSON
{
|
||
"config": {
|
||
"ckpt": "ml/runs/base_dag2_dag1/best.pt",
|
||
"samples": "ml/data/samples",
|
||
"tasks": null,
|
||
"split": "val",
|
||
"n": 64,
|
||
"K": 8,
|
||
"commit": 2,
|
||
"fb": 2.2,
|
||
"ki": 0.0,
|
||
"fb_scale": 0.0,
|
||
"fb_near": 0.2,
|
||
"fb_fallback": 0.0,
|
||
"kdecay": 0.9,
|
||
"kimax": 0.5,
|
||
"fb_every": 1,
|
||
"safety_try": 0,
|
||
"sel_w": 0.1,
|
||
"dls_safety": 1,
|
||
"fb_clear": 0.0,
|
||
"fb_cmin": 0.3,
|
||
"sub_refresh": 0,
|
||
"fb_block_gain": 0.5,
|
||
"dump": "ml/runs/round44_tieprobe64_rows.json",
|
||
"out": "docs/experiment_round44_tieprobe64.json",
|
||
"trace": null,
|
||
"select": "endpoint",
|
||
"replan": "empty",
|
||
"safety": 1,
|
||
"partial": 1,
|
||
"inc_cap": 1,
|
||
"gap_margin": 0.0,
|
||
"cert_batch": 1,
|
||
"steps": 1,
|
||
"z_scale": 1.0,
|
||
"len_w": 0.0,
|
||
"tie": 1,
|
||
"backoff": 0,
|
||
"max_sec": null,
|
||
"seed": 0,
|
||
"dls": "analytic",
|
||
"device": "cuda"
|
||
},
|
||
"batch_size": 256,
|
||
"batch_seed_rule": "same seed per batch",
|
||
"checkpoint_sha256": "f7f686fd8ee92fb6948fbcc4ee954636065ad87bb54f661c4f8f575bc16c2ce1",
|
||
"tasks_sha256": "6185a312dbae3bf5b361c4b9ff3aa11b7329d9723cff2be9b9157563d6b6ec1c",
|
||
"kin_fp": "2ebcbf17",
|
||
"act_scale": 1.0,
|
||
"metrics": {
|
||
"n": 64,
|
||
"n_ref": 64,
|
||
"success": 0.40625,
|
||
"collision_free_success": 0.40625,
|
||
"time_ok_1p3": 0.234375,
|
||
"collision_rate": 0.0,
|
||
"collision_steps_mean": 0.0,
|
||
"travel_ratio_median": 2.358931303024292,
|
||
"time_mean_s": 2.3323075771331787,
|
||
"brakes_mean": 18.71875,
|
||
"start_fail": 0.0,
|
||
"goal_ratio_2s": 0.33309799432754517,
|
||
"err_fail_med": 229.02410888671875,
|
||
"err_success_med": 0.22094391286373138,
|
||
"dls_share": 0.515625,
|
||
"time_budget": "teacher×2(≥3 s)",
|
||
"speed_ratio_p50": 1.0000016689300537,
|
||
"speed_ratio_p95": 1.0000091791152954,
|
||
"speed_ratio_max": 1.0000101327896118,
|
||
"over_vmax_rate": 0.0,
|
||
"accel_ratio_p95": 1.0000016689300537,
|
||
"accel_ratio_max": 1.0000101327896118,
|
||
"reverse_p50": 0.25150150060653687,
|
||
"reverse_mean": 0.25861242413520813,
|
||
"reverse_moving_p50": 0.318452388048172,
|
||
"moving_share_p50": 0.9090909361839294,
|
||
"brake_rate_mean": 0.10696429014205933,
|
||
"brake_rate_p50": 0.017142856493592262,
|
||
"brake_rate_p90": 0.33142855763435364,
|
||
"stuck_share": 0.21875,
|
||
"n_plans_mean": 175.0,
|
||
"success_reachable": 0.4262295067310333,
|
||
"by_cat": {
|
||
"A": [
|
||
28,
|
||
14
|
||
],
|
||
"B": [
|
||
21,
|
||
6
|
||
],
|
||
"C": [
|
||
9,
|
||
4
|
||
],
|
||
"D": [
|
||
3,
|
||
0
|
||
],
|
||
"Dr": [
|
||
3,
|
||
2
|
||
]
|
||
},
|
||
"tie_opp_rate_mean": 0.7484821081161499,
|
||
"tie_changed_rate_mean": 0.07437500357627869,
|
||
"tie_opp_task_share": 1.0
|
||
},
|
||
"simulation_wall_s": 78.19308080000337
|
||
}
|