1{2 "system": "p4v2_gr_s2",3 "split": "test_depth",4 "checkpoint": "experiments/runs/p4v2_gr_seed02_20261030001749/checkpoint",5 "checkpoint_sha256": "13e35362d3e2e4c169299b94379e544e3b64dde1d8cb8d7b0c5c32346d279037",6 "config_sha256": "0d380d727de6614b17db8524bf8e936a06211456dbd6fc7d3324b5d06e75fe70",7 "code_commit": "unknown",8 "device": "mps",9 "precision": "fp32",10 "evaluated_at": "2026-10-30T04:06:11-0400",11 "data_sha256": "06620e1cf60b7250aac71f4fb5ee74e3705f4b9aab102e22b5c5bdbed76c4ae2",12 "exact_match": {13 "budget0": {14 "accuracy": 0.312,15 "n": 1000,16 "by_family": {17 "arith": 0.0,18 "bind": 0.02,19 "fsm": 0.425,20 "excl": 0.81,21 "sched": 0.30522 },23 "edits_per_token": 0.024 },25 "forced_b2": {26 "accuracy": 0.33,27 "n": 1000,28 "by_family": {29 "arith": 0.0,30 "bind": 0.035,31 "fsm": 0.455,32 "excl": 0.795,33 "sched": 0.36534 },35 "edits_per_token": 2.036 }37 },38 "seconds": 12.039}