diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,3633 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.08016032064128256,
+  "eval_steps": 500,
+  "global_step": 300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completion_length": 94.5250015258789,
+      "epoch": 0.0002672010688042752,
+      "grad_norm": 8.340370178222656,
+      "kl": 0.0,
+      "learning_rate": 9.997327632282202e-07,
+      "loss": 0.0,
+      "reward": 0.43238671123981476,
+      "reward_std": 0.10043386369943619,
+      "rewards/iou_timestamp_reward": 0.43238671123981476,
+      "step": 1
+    },
+    {
+      "completion_length": 90.17499923706055,
+      "epoch": 0.0005344021376085504,
+      "grad_norm": 15.389718055725098,
+      "kl": 0.0009441375732421875,
+      "learning_rate": 9.994655264564402e-07,
+      "loss": 0.0,
+      "reward": 0.49521203339099884,
+      "reward_std": 0.1284736730158329,
+      "rewards/iou_timestamp_reward": 0.49521203339099884,
+      "step": 2
+    },
+    {
+      "completion_length": 94.7750015258789,
+      "epoch": 0.0008016032064128256,
+      "grad_norm": 14.409242630004883,
+      "kl": 0.001331329345703125,
+      "learning_rate": 9.991982896846605e-07,
+      "loss": 0.0001,
+      "reward": 0.5351988673210144,
+      "reward_std": 0.16380538046360016,
+      "rewards/iou_timestamp_reward": 0.5351988673210144,
+      "step": 3
+    },
+    {
+      "completion_length": 92.67499923706055,
+      "epoch": 0.0010688042752171009,
+      "grad_norm": 12.140655517578125,
+      "kl": 0.001255035400390625,
+      "learning_rate": 9.989310529128807e-07,
+      "loss": 0.0001,
+      "reward": 0.3954579681158066,
+      "reward_std": 0.1440194696187973,
+      "rewards/iou_timestamp_reward": 0.3954579681158066,
+      "step": 4
+    },
+    {
+      "completion_length": 93.0250015258789,
+      "epoch": 0.0013360053440213762,
+      "grad_norm": 11.99720573425293,
+      "kl": 0.00121307373046875,
+      "learning_rate": 9.98663816141101e-07,
+      "loss": 0.0,
+      "reward": 0.5536792427301407,
+      "reward_std": 0.1233767680823803,
+      "rewards/iou_timestamp_reward": 0.5536792427301407,
+      "step": 5
+    },
+    {
+      "completion_length": 95.32500076293945,
+      "epoch": 0.0016032064128256513,
+      "grad_norm": 31.318050384521484,
+      "kl": 0.001331329345703125,
+      "learning_rate": 9.983965793693212e-07,
+      "loss": 0.0001,
+      "reward": 0.5504237711429596,
+      "reward_std": 0.17604028433561325,
+      "rewards/iou_timestamp_reward": 0.5504237711429596,
+      "step": 6
+    },
+    {
+      "completion_length": 96.47500228881836,
+      "epoch": 0.0018704074816299266,
+      "grad_norm": 31.36016273498535,
+      "kl": 0.001834869384765625,
+      "learning_rate": 9.981293425975412e-07,
+      "loss": 0.0001,
+      "reward": 0.3581559360027313,
+      "reward_std": 0.07435427978634834,
+      "rewards/iou_timestamp_reward": 0.3581559360027313,
+      "step": 7
+    },
+    {
+      "completion_length": 90.57500076293945,
+      "epoch": 0.0021376085504342017,
+      "grad_norm": 12.298426628112793,
+      "kl": 0.00214385986328125,
+      "learning_rate": 9.978621058257615e-07,
+      "loss": 0.0001,
+      "reward": 0.5335834175348282,
+      "reward_std": 0.09592974931001663,
+      "rewards/iou_timestamp_reward": 0.5335834175348282,
+      "step": 8
+    },
+    {
+      "completion_length": 84.61250305175781,
+      "epoch": 0.002404809619238477,
+      "grad_norm": 8.306353569030762,
+      "kl": 0.00254058837890625,
+      "learning_rate": 9.975948690539817e-07,
+      "loss": 0.0001,
+      "reward": 0.4508528411388397,
+      "reward_std": 0.1074596531689167,
+      "rewards/iou_timestamp_reward": 0.4508528411388397,
+      "step": 9
+    },
+    {
+      "completion_length": 90.4000015258789,
+      "epoch": 0.0026720106880427524,
+      "grad_norm": 9.342815399169922,
+      "kl": 0.002593994140625,
+      "learning_rate": 9.97327632282202e-07,
+      "loss": 0.0001,
+      "reward": 0.567167192697525,
+      "reward_std": 0.13745272532105446,
+      "rewards/iou_timestamp_reward": 0.567167192697525,
+      "step": 10
+    },
+    {
+      "completion_length": 87.07500076293945,
+      "epoch": 0.0029392117568470275,
+      "grad_norm": 22.01643943786621,
+      "kl": 0.00392913818359375,
+      "learning_rate": 9.970603955104222e-07,
+      "loss": 0.0002,
+      "reward": 0.4692964553833008,
+      "reward_std": 0.0816560797393322,
+      "rewards/iou_timestamp_reward": 0.4692964553833008,
+      "step": 11
+    },
+    {
+      "completion_length": 86.0625,
+      "epoch": 0.0032064128256513026,
+      "grad_norm": 19.23089599609375,
+      "kl": 0.00360870361328125,
+      "learning_rate": 9.967931587386423e-07,
+      "loss": 0.0001,
+      "reward": 0.35264405608177185,
+      "reward_std": 0.0850561149418354,
+      "rewards/iou_timestamp_reward": 0.35264405608177185,
+      "step": 12
+    },
+    {
+      "completion_length": 102.80000305175781,
+      "epoch": 0.0034736138944555777,
+      "grad_norm": 19.64198875427246,
+      "kl": 0.0048370361328125,
+      "learning_rate": 9.965259219668625e-07,
+      "loss": 0.0002,
+      "reward": 0.48492929339408875,
+      "reward_std": 0.13556858897209167,
+      "rewards/iou_timestamp_reward": 0.48492929339408875,
+      "step": 13
+    },
+    {
+      "completion_length": 85.86249923706055,
+      "epoch": 0.0037408149632598532,
+      "grad_norm": 9.946146011352539,
+      "kl": 0.0070953369140625,
+      "learning_rate": 9.962586851950827e-07,
+      "loss": 0.0003,
+      "reward": 0.5720771998167038,
+      "reward_std": 0.12591452151536942,
+      "rewards/iou_timestamp_reward": 0.5720771998167038,
+      "step": 14
+    },
+    {
+      "completion_length": 90.0625,
+      "epoch": 0.004008016032064128,
+      "grad_norm": 13.546260833740234,
+      "kl": 0.006072998046875,
+      "learning_rate": 9.95991448423303e-07,
+      "loss": 0.0002,
+      "reward": 0.5297239720821381,
+      "reward_std": 0.10710490867495537,
+      "rewards/iou_timestamp_reward": 0.5297239720821381,
+      "step": 15
+    },
+    {
+      "completion_length": 83.01250076293945,
+      "epoch": 0.0042752171008684035,
+      "grad_norm": 8.076241493225098,
+      "kl": 0.0060882568359375,
+      "learning_rate": 9.957242116515232e-07,
+      "loss": 0.0002,
+      "reward": 0.29969294369220734,
+      "reward_std": 0.09227840602397919,
+      "rewards/iou_timestamp_reward": 0.29969294369220734,
+      "step": 16
+    },
+    {
+      "completion_length": 75.48750305175781,
+      "epoch": 0.004542418169672679,
+      "grad_norm": 7.656135559082031,
+      "kl": 0.0081329345703125,
+      "learning_rate": 9.954569748797435e-07,
+      "loss": 0.0003,
+      "reward": 0.5397222191095352,
+      "reward_std": 0.12004014477133751,
+      "rewards/iou_timestamp_reward": 0.5397222191095352,
+      "step": 17
+    },
+    {
+      "completion_length": 89.41250228881836,
+      "epoch": 0.004809619238476954,
+      "grad_norm": 15.919727325439453,
+      "kl": 0.0092315673828125,
+      "learning_rate": 9.951897381079635e-07,
+      "loss": 0.0004,
+      "reward": 0.4049786627292633,
+      "reward_std": 0.09774354100227356,
+      "rewards/iou_timestamp_reward": 0.4049786627292633,
+      "step": 18
+    },
+    {
+      "completion_length": 84.35000228881836,
+      "epoch": 0.005076820307281229,
+      "grad_norm": 7.11904764175415,
+      "kl": 0.008270263671875,
+      "learning_rate": 9.949225013361838e-07,
+      "loss": 0.0003,
+      "reward": 0.36138640344142914,
+      "reward_std": 0.09036249667406082,
+      "rewards/iou_timestamp_reward": 0.36138640344142914,
+      "step": 19
+    },
+    {
+      "completion_length": 84.35000228881836,
+      "epoch": 0.005344021376085505,
+      "grad_norm": 31.289043426513672,
+      "kl": 0.0091552734375,
+      "learning_rate": 9.94655264564404e-07,
+      "loss": 0.0004,
+      "reward": 0.4811190813779831,
+      "reward_std": 0.09721268340945244,
+      "rewards/iou_timestamp_reward": 0.4811190813779831,
+      "step": 20
+    },
+    {
+      "completion_length": 81.3375015258789,
+      "epoch": 0.0056112224448897794,
+      "grad_norm": 25.466672897338867,
+      "kl": 0.010284423828125,
+      "learning_rate": 9.943880277926243e-07,
+      "loss": 0.0004,
+      "reward": 0.5279562920331955,
+      "reward_std": 0.05652128346264362,
+      "rewards/iou_timestamp_reward": 0.5279562920331955,
+      "step": 21
+    },
+    {
+      "completion_length": 82.80000305175781,
+      "epoch": 0.005878423513694055,
+      "grad_norm": 19.14952278137207,
+      "kl": 0.009918212890625,
+      "learning_rate": 9.941207910208445e-07,
+      "loss": 0.0004,
+      "reward": 0.4550338685512543,
+      "reward_std": 0.08861872926354408,
+      "rewards/iou_timestamp_reward": 0.4550338685512543,
+      "step": 22
+    },
+    {
+      "completion_length": 82.03750228881836,
+      "epoch": 0.00614562458249833,
+      "grad_norm": 20.35788345336914,
+      "kl": 0.010284423828125,
+      "learning_rate": 9.938535542490645e-07,
+      "loss": 0.0004,
+      "reward": 0.5463799238204956,
+      "reward_std": 0.1116589643061161,
+      "rewards/iou_timestamp_reward": 0.5463799238204956,
+      "step": 23
+    },
+    {
+      "completion_length": 82.80000305175781,
+      "epoch": 0.006412825651302605,
+      "grad_norm": 14.981146812438965,
+      "kl": 0.0108642578125,
+      "learning_rate": 9.935863174772848e-07,
+      "loss": 0.0004,
+      "reward": 0.4238290339708328,
+      "reward_std": 0.0759831890463829,
+      "rewards/iou_timestamp_reward": 0.4238290339708328,
+      "step": 24
+    },
+    {
+      "completion_length": 81.72500228881836,
+      "epoch": 0.006680026720106881,
+      "grad_norm": 15.833313941955566,
+      "kl": 0.011749267578125,
+      "learning_rate": 9.93319080705505e-07,
+      "loss": 0.0005,
+      "reward": 0.3965422213077545,
+      "reward_std": 0.07669836282730103,
+      "rewards/iou_timestamp_reward": 0.3965422213077545,
+      "step": 25
+    },
+    {
+      "completion_length": 93.2750015258789,
+      "epoch": 0.006947227788911155,
+      "grad_norm": 45.69765090942383,
+      "kl": 0.011199951171875,
+      "learning_rate": 9.930518439337253e-07,
+      "loss": 0.0004,
+      "reward": 0.548987627029419,
+      "reward_std": 0.1403028443455696,
+      "rewards/iou_timestamp_reward": 0.548987627029419,
+      "step": 26
+    },
+    {
+      "completion_length": 81.04999923706055,
+      "epoch": 0.007214428857715431,
+      "grad_norm": 10.686071395874023,
+      "kl": 0.0106201171875,
+      "learning_rate": 9.927846071619455e-07,
+      "loss": 0.0004,
+      "reward": 0.3573782742023468,
+      "reward_std": 0.11069009453058243,
+      "rewards/iou_timestamp_reward": 0.3573782742023468,
+      "step": 27
+    },
+    {
+      "completion_length": 78.625,
+      "epoch": 0.0074816299265197065,
+      "grad_norm": 8.17104721069336,
+      "kl": 0.012176513671875,
+      "learning_rate": 9.925173703901655e-07,
+      "loss": 0.0005,
+      "reward": 0.34273259341716766,
+      "reward_std": 0.08138729631900787,
+      "rewards/iou_timestamp_reward": 0.34273259341716766,
+      "step": 28
+    },
+    {
+      "completion_length": 78.26250076293945,
+      "epoch": 0.007748830995323981,
+      "grad_norm": 31.360939025878906,
+      "kl": 0.013702392578125,
+      "learning_rate": 9.922501336183858e-07,
+      "loss": 0.0005,
+      "reward": 0.45369820296764374,
+      "reward_std": 0.10742013901472092,
+      "rewards/iou_timestamp_reward": 0.45369820296764374,
+      "step": 29
+    },
+    {
+      "completion_length": 75.22500228881836,
+      "epoch": 0.008016032064128256,
+      "grad_norm": 9.121045112609863,
+      "kl": 0.01019287109375,
+      "learning_rate": 9.91982896846606e-07,
+      "loss": 0.0004,
+      "reward": 0.3313401788473129,
+      "reward_std": 0.13614049553871155,
+      "rewards/iou_timestamp_reward": 0.3313401788473129,
+      "step": 30
+    },
+    {
+      "completion_length": 89.67500305175781,
+      "epoch": 0.008283233132932532,
+      "grad_norm": 19.901443481445312,
+      "kl": 0.013580322265625,
+      "learning_rate": 9.917156600748263e-07,
+      "loss": 0.0005,
+      "reward": 0.5152908265590668,
+      "reward_std": 0.09363758563995361,
+      "rewards/iou_timestamp_reward": 0.5152908265590668,
+      "step": 31
+    },
+    {
+      "completion_length": 87.0250015258789,
+      "epoch": 0.008550434201736807,
+      "grad_norm": 18.844900131225586,
+      "kl": 0.01513671875,
+      "learning_rate": 9.914484233030465e-07,
+      "loss": 0.0006,
+      "reward": 0.47241438925266266,
+      "reward_std": 0.11852370575070381,
+      "rewards/iou_timestamp_reward": 0.47241438925266266,
+      "step": 32
+    },
+    {
+      "completion_length": 77.43750381469727,
+      "epoch": 0.008817635270541082,
+      "grad_norm": 10.22252082824707,
+      "kl": 0.01824951171875,
+      "learning_rate": 9.911811865312668e-07,
+      "loss": 0.0007,
+      "reward": 0.44154876470565796,
+      "reward_std": 0.08548092842102051,
+      "rewards/iou_timestamp_reward": 0.44154876470565796,
+      "step": 33
+    },
+    {
+      "completion_length": 82.98750305175781,
+      "epoch": 0.009084836339345358,
+      "grad_norm": 17.666988372802734,
+      "kl": 0.01806640625,
+      "learning_rate": 9.909139497594868e-07,
+      "loss": 0.0007,
+      "reward": 0.3770228624343872,
+      "reward_std": 0.09334432706236839,
+      "rewards/iou_timestamp_reward": 0.3770228624343872,
+      "step": 34
+    },
+    {
+      "completion_length": 80.98750305175781,
+      "epoch": 0.009352037408149633,
+      "grad_norm": 13.110654830932617,
+      "kl": 0.016082763671875,
+      "learning_rate": 9.90646712987707e-07,
+      "loss": 0.0006,
+      "reward": 0.4078673869371414,
+      "reward_std": 0.059798676520586014,
+      "rewards/iou_timestamp_reward": 0.4078673869371414,
+      "step": 35
+    },
+    {
+      "completion_length": 92.25000381469727,
+      "epoch": 0.009619238476953907,
+      "grad_norm": 8.614245414733887,
+      "kl": 0.01861572265625,
+      "learning_rate": 9.903794762159273e-07,
+      "loss": 0.0007,
+      "reward": 0.35014285147190094,
+      "reward_std": 0.0985935740172863,
+      "rewards/iou_timestamp_reward": 0.35014285147190094,
+      "step": 36
+    },
+    {
+      "completion_length": 80.56250381469727,
+      "epoch": 0.009886439545758184,
+      "grad_norm": 10.115817070007324,
+      "kl": 0.01837158203125,
+      "learning_rate": 9.901122394441475e-07,
+      "loss": 0.0007,
+      "reward": 0.40472397208213806,
+      "reward_std": 0.12752560898661613,
+      "rewards/iou_timestamp_reward": 0.40472397208213806,
+      "step": 37
+    },
+    {
+      "completion_length": 81.17500305175781,
+      "epoch": 0.010153640614562458,
+      "grad_norm": 39.667320251464844,
+      "kl": 0.015625,
+      "learning_rate": 9.898450026723678e-07,
+      "loss": 0.0006,
+      "reward": 0.46072031557559967,
+      "reward_std": 0.09041710384190083,
+      "rewards/iou_timestamp_reward": 0.46072031557559967,
+      "step": 38
+    },
+    {
+      "completion_length": 81.61250305175781,
+      "epoch": 0.010420841683366733,
+      "grad_norm": 13.683932304382324,
+      "kl": 0.016387939453125,
+      "learning_rate": 9.895777659005878e-07,
+      "loss": 0.0007,
+      "reward": 0.543788880109787,
+      "reward_std": 0.10228582844138145,
+      "rewards/iou_timestamp_reward": 0.543788880109787,
+      "step": 39
+    },
+    {
+      "completion_length": 81.35000228881836,
+      "epoch": 0.01068804275217101,
+      "grad_norm": 12.15194034576416,
+      "kl": 0.017791748046875,
+      "learning_rate": 9.89310529128808e-07,
+      "loss": 0.0007,
+      "reward": 0.4217175394296646,
+      "reward_std": 0.12605417147278786,
+      "rewards/iou_timestamp_reward": 0.4217175394296646,
+      "step": 40
+    },
+    {
+      "completion_length": 81.98750305175781,
+      "epoch": 0.010955243820975284,
+      "grad_norm": 121.88482666015625,
+      "kl": 0.01617431640625,
+      "learning_rate": 9.890432923570283e-07,
+      "loss": 0.0006,
+      "reward": 0.36619891971349716,
+      "reward_std": 0.10043040104210377,
+      "rewards/iou_timestamp_reward": 0.36619891971349716,
+      "step": 41
+    },
+    {
+      "completion_length": 76.97500228881836,
+      "epoch": 0.011222444889779559,
+      "grad_norm": 14.41795825958252,
+      "kl": 0.0224609375,
+      "learning_rate": 9.887760555852485e-07,
+      "loss": 0.0009,
+      "reward": 0.3520086780190468,
+      "reward_std": 0.09863714687526226,
+      "rewards/iou_timestamp_reward": 0.3520086780190468,
+      "step": 42
+    },
+    {
+      "completion_length": 84.72500228881836,
+      "epoch": 0.011489645958583834,
+      "grad_norm": 14.718040466308594,
+      "kl": 0.01531982421875,
+      "learning_rate": 9.885088188134688e-07,
+      "loss": 0.0006,
+      "reward": 0.5319889932870865,
+      "reward_std": 0.11806132644414902,
+      "rewards/iou_timestamp_reward": 0.5319889932870865,
+      "step": 43
+    },
+    {
+      "completion_length": 83.61250305175781,
+      "epoch": 0.01175684702738811,
+      "grad_norm": 18.669189453125,
+      "kl": 0.0208740234375,
+      "learning_rate": 9.882415820416888e-07,
+      "loss": 0.0008,
+      "reward": 0.5544195771217346,
+      "reward_std": 0.11862267926335335,
+      "rewards/iou_timestamp_reward": 0.5544195771217346,
+      "step": 44
+    },
+    {
+      "completion_length": 82.10000228881836,
+      "epoch": 0.012024048096192385,
+      "grad_norm": 13.057522773742676,
+      "kl": 0.01544189453125,
+      "learning_rate": 9.87974345269909e-07,
+      "loss": 0.0006,
+      "reward": 0.4692879766225815,
+      "reward_std": 0.08527671545743942,
+      "rewards/iou_timestamp_reward": 0.4692879766225815,
+      "step": 45
+    },
+    {
+      "completion_length": 81.8125,
+      "epoch": 0.01229124916499666,
+      "grad_norm": 62.2025260925293,
+      "kl": 0.021392822265625,
+      "learning_rate": 9.877071084981293e-07,
+      "loss": 0.0009,
+      "reward": 0.4865482598543167,
+      "reward_std": 0.11775477230548859,
+      "rewards/iou_timestamp_reward": 0.4865482598543167,
+      "step": 46
+    },
+    {
+      "completion_length": 81.88750457763672,
+      "epoch": 0.012558450233800936,
+      "grad_norm": 11.27963638305664,
+      "kl": 0.02313232421875,
+      "learning_rate": 9.874398717263496e-07,
+      "loss": 0.0009,
+      "reward": 0.4721195548772812,
+      "reward_std": 0.0720259640365839,
+      "rewards/iou_timestamp_reward": 0.4721195548772812,
+      "step": 47
+    },
+    {
+      "completion_length": 81.82500457763672,
+      "epoch": 0.01282565130260521,
+      "grad_norm": 9.9296875,
+      "kl": 0.016571044921875,
+      "learning_rate": 9.871726349545698e-07,
+      "loss": 0.0007,
+      "reward": 0.48466672003269196,
+      "reward_std": 0.15841123461723328,
+      "rewards/iou_timestamp_reward": 0.48466672003269196,
+      "step": 48
+    },
+    {
+      "completion_length": 79.66250228881836,
+      "epoch": 0.013092852371409485,
+      "grad_norm": 10.52200698852539,
+      "kl": 0.01617431640625,
+      "learning_rate": 9.8690539818279e-07,
+      "loss": 0.0006,
+      "reward": 0.3634609803557396,
+      "reward_std": 0.13740423694252968,
+      "rewards/iou_timestamp_reward": 0.3634609803557396,
+      "step": 49
+    },
+    {
+      "completion_length": 82.2750015258789,
+      "epoch": 0.013360053440213761,
+      "grad_norm": 19.474987030029297,
+      "kl": 0.0179443359375,
+      "learning_rate": 9.8663816141101e-07,
+      "loss": 0.0007,
+      "reward": 0.5607456266880035,
+      "reward_std": 0.11081211641430855,
+      "rewards/iou_timestamp_reward": 0.5607456266880035,
+      "step": 50
+    },
+    {
+      "completion_length": 91.9625015258789,
+      "epoch": 0.013627254509018036,
+      "grad_norm": 8.144207000732422,
+      "kl": 0.0185546875,
+      "learning_rate": 9.863709246392303e-07,
+      "loss": 0.0007,
+      "reward": 0.5219191014766693,
+      "reward_std": 0.12948721647262573,
+      "rewards/iou_timestamp_reward": 0.5219191014766693,
+      "step": 51
+    },
+    {
+      "completion_length": 82.75,
+      "epoch": 0.01389445557782231,
+      "grad_norm": 20.344608306884766,
+      "kl": 0.02337646484375,
+      "learning_rate": 9.861036878674504e-07,
+      "loss": 0.0009,
+      "reward": 0.4568335562944412,
+      "reward_std": 0.06874274648725986,
+      "rewards/iou_timestamp_reward": 0.4568335562944412,
+      "step": 52
+    },
+    {
+      "completion_length": 86.7750015258789,
+      "epoch": 0.014161656646626587,
+      "grad_norm": 21.093334197998047,
+      "kl": 0.01800537109375,
+      "learning_rate": 9.858364510956706e-07,
+      "loss": 0.0007,
+      "reward": 0.5438729226589203,
+      "reward_std": 0.14354269951581955,
+      "rewards/iou_timestamp_reward": 0.5438729226589203,
+      "step": 53
+    },
+    {
+      "completion_length": 81.17500305175781,
+      "epoch": 0.014428857715430862,
+      "grad_norm": 13.528135299682617,
+      "kl": 0.02130126953125,
+      "learning_rate": 9.855692143238909e-07,
+      "loss": 0.0009,
+      "reward": 0.43492650985717773,
+      "reward_std": 0.08151087537407875,
+      "rewards/iou_timestamp_reward": 0.43492650985717773,
+      "step": 54
+    },
+    {
+      "completion_length": 80.125,
+      "epoch": 0.014696058784235137,
+      "grad_norm": 28.23933219909668,
+      "kl": 0.0201416015625,
+      "learning_rate": 9.85301977552111e-07,
+      "loss": 0.0008,
+      "reward": 0.5359328091144562,
+      "reward_std": 0.11250880360603333,
+      "rewards/iou_timestamp_reward": 0.5359328091144562,
+      "step": 55
+    },
+    {
+      "completion_length": 82.36249923706055,
+      "epoch": 0.014963259853039413,
+      "grad_norm": 28.83526611328125,
+      "kl": 0.0228271484375,
+      "learning_rate": 9.850347407803313e-07,
+      "loss": 0.0009,
+      "reward": 0.41243691742420197,
+      "reward_std": 0.12655200064182281,
+      "rewards/iou_timestamp_reward": 0.41243691742420197,
+      "step": 56
+    },
+    {
+      "completion_length": 82.375,
+      "epoch": 0.015230460921843688,
+      "grad_norm": 11.394326210021973,
+      "kl": 0.023193359375,
+      "learning_rate": 9.847675040085516e-07,
+      "loss": 0.0009,
+      "reward": 0.44033171236515045,
+      "reward_std": 0.06593804247677326,
+      "rewards/iou_timestamp_reward": 0.44033171236515045,
+      "step": 57
+    },
+    {
+      "completion_length": 77.88750076293945,
+      "epoch": 0.015497661990647962,
+      "grad_norm": 40.8941535949707,
+      "kl": 0.025146484375,
+      "learning_rate": 9.845002672367716e-07,
+      "loss": 0.001,
+      "reward": 0.3637496903538704,
+      "reward_std": 0.040244522504508495,
+      "rewards/iou_timestamp_reward": 0.3637496903538704,
+      "step": 58
+    },
+    {
+      "completion_length": 82.3375015258789,
+      "epoch": 0.01576486305945224,
+      "grad_norm": 14.911782264709473,
+      "kl": 0.02606201171875,
+      "learning_rate": 9.842330304649919e-07,
+      "loss": 0.001,
+      "reward": 0.5161260962486267,
+      "reward_std": 0.07627514749765396,
+      "rewards/iou_timestamp_reward": 0.5161260962486267,
+      "step": 59
+    },
+    {
+      "completion_length": 80.5374984741211,
+      "epoch": 0.01603206412825651,
+      "grad_norm": 17.467588424682617,
+      "kl": 0.02093505859375,
+      "learning_rate": 9.839657936932121e-07,
+      "loss": 0.0008,
+      "reward": 0.3705238401889801,
+      "reward_std": 0.06593567878007889,
+      "rewards/iou_timestamp_reward": 0.3705238401889801,
+      "step": 60
+    },
+    {
+      "completion_length": 77.6624984741211,
+      "epoch": 0.016299265197060788,
+      "grad_norm": 21.349925994873047,
+      "kl": 0.0181884765625,
+      "learning_rate": 9.836985569214324e-07,
+      "loss": 0.0007,
+      "reward": 0.662094235420227,
+      "reward_std": 0.09039272367954254,
+      "rewards/iou_timestamp_reward": 0.662094235420227,
+      "step": 61
+    },
+    {
+      "completion_length": 84.00000381469727,
+      "epoch": 0.016566466265865065,
+      "grad_norm": 7.912447929382324,
+      "kl": 0.02349853515625,
+      "learning_rate": 9.834313201496526e-07,
+      "loss": 0.0009,
+      "reward": 0.55202616751194,
+      "reward_std": 0.07802803814411163,
+      "rewards/iou_timestamp_reward": 0.55202616751194,
+      "step": 62
+    },
+    {
+      "completion_length": 79.85000228881836,
+      "epoch": 0.016833667334669337,
+      "grad_norm": 18.316444396972656,
+      "kl": 0.03076171875,
+      "learning_rate": 9.831640833778726e-07,
+      "loss": 0.0012,
+      "reward": 0.49671244621276855,
+      "reward_std": 0.07321073487401009,
+      "rewards/iou_timestamp_reward": 0.49671244621276855,
+      "step": 63
+    },
+    {
+      "completion_length": 75.6500015258789,
+      "epoch": 0.017100868403473614,
+      "grad_norm": 32.6612663269043,
+      "kl": 0.0294189453125,
+      "learning_rate": 9.828968466060929e-07,
+      "loss": 0.0012,
+      "reward": 0.4359857589006424,
+      "reward_std": 0.1010651346296072,
+      "rewards/iou_timestamp_reward": 0.4359857589006424,
+      "step": 64
+    },
+    {
+      "completion_length": 84.8125,
+      "epoch": 0.01736806947227789,
+      "grad_norm": 19.969472885131836,
+      "kl": 0.0228271484375,
+      "learning_rate": 9.826296098343131e-07,
+      "loss": 0.0009,
+      "reward": 0.36607126891613007,
+      "reward_std": 0.08596941456198692,
+      "rewards/iou_timestamp_reward": 0.36607126891613007,
+      "step": 65
+    },
+    {
+      "completion_length": 82.43750381469727,
+      "epoch": 0.017635270541082163,
+      "grad_norm": 10.52409553527832,
+      "kl": 0.0196533203125,
+      "learning_rate": 9.823623730625334e-07,
+      "loss": 0.0008,
+      "reward": 0.4937574118375778,
+      "reward_std": 0.09046215564012527,
+      "rewards/iou_timestamp_reward": 0.4937574118375778,
+      "step": 66
+    },
+    {
+      "completion_length": 82.68750381469727,
+      "epoch": 0.01790247160988644,
+      "grad_norm": 13.098502159118652,
+      "kl": 0.02655029296875,
+      "learning_rate": 9.820951362907536e-07,
+      "loss": 0.0011,
+      "reward": 0.5216490924358368,
+      "reward_std": 0.08065400645136833,
+      "rewards/iou_timestamp_reward": 0.5216490924358368,
+      "step": 67
+    },
+    {
+      "completion_length": 80.88750076293945,
+      "epoch": 0.018169672678690716,
+      "grad_norm": 7.5821452140808105,
+      "kl": 0.02789306640625,
+      "learning_rate": 9.818278995189736e-07,
+      "loss": 0.0011,
+      "reward": 0.5076297074556351,
+      "reward_std": 0.08006440289318562,
+      "rewards/iou_timestamp_reward": 0.5076297074556351,
+      "step": 68
+    },
+    {
+      "completion_length": 82.78750228881836,
+      "epoch": 0.01843687374749499,
+      "grad_norm": 20.07110595703125,
+      "kl": 0.0191650390625,
+      "learning_rate": 9.815606627471939e-07,
+      "loss": 0.0008,
+      "reward": 0.32277095317840576,
+      "reward_std": 0.05622895993292332,
+      "rewards/iou_timestamp_reward": 0.32277095317840576,
+      "step": 69
+    },
+    {
+      "completion_length": 83.57500457763672,
+      "epoch": 0.018704074816299265,
+      "grad_norm": 58.287322998046875,
+      "kl": 0.02606201171875,
+      "learning_rate": 9.812934259754141e-07,
+      "loss": 0.001,
+      "reward": 0.47261086106300354,
+      "reward_std": 0.1284152865409851,
+      "rewards/iou_timestamp_reward": 0.47261086106300354,
+      "step": 70
+    },
+    {
+      "completion_length": 76.6875,
+      "epoch": 0.018971275885103542,
+      "grad_norm": 15.780878067016602,
+      "kl": 0.026611328125,
+      "learning_rate": 9.810261892036344e-07,
+      "loss": 0.0011,
+      "reward": 0.4276278167963028,
+      "reward_std": 0.10460241883993149,
+      "rewards/iou_timestamp_reward": 0.4276278167963028,
+      "step": 71
+    },
+    {
+      "completion_length": 76.17500305175781,
+      "epoch": 0.019238476953907815,
+      "grad_norm": 354.0404968261719,
+      "kl": 0.02935791015625,
+      "learning_rate": 9.807589524318546e-07,
+      "loss": 0.0012,
+      "reward": 0.6388321816921234,
+      "reward_std": 0.10210368782281876,
+      "rewards/iou_timestamp_reward": 0.6388321816921234,
+      "step": 72
+    },
+    {
+      "completion_length": 74.32500076293945,
+      "epoch": 0.01950567802271209,
+      "grad_norm": 8.050577163696289,
+      "kl": 0.02825927734375,
+      "learning_rate": 9.804917156600749e-07,
+      "loss": 0.0011,
+      "reward": 0.3472304940223694,
+      "reward_std": 0.16149252653121948,
+      "rewards/iou_timestamp_reward": 0.3472304940223694,
+      "step": 73
+    },
+    {
+      "completion_length": 83.6624984741211,
+      "epoch": 0.019772879091516368,
+      "grad_norm": 16.733179092407227,
+      "kl": 0.0260009765625,
+      "learning_rate": 9.80224478888295e-07,
+      "loss": 0.001,
+      "reward": 0.6152356266975403,
+      "reward_std": 0.09945449233055115,
+      "rewards/iou_timestamp_reward": 0.6152356266975403,
+      "step": 74
+    },
+    {
+      "completion_length": 82.03750228881836,
+      "epoch": 0.02004008016032064,
+      "grad_norm": 11.940522193908691,
+      "kl": 0.038330078125,
+      "learning_rate": 9.799572421165151e-07,
+      "loss": 0.0015,
+      "reward": 0.6018393933773041,
+      "reward_std": 0.10049806535243988,
+      "rewards/iou_timestamp_reward": 0.6018393933773041,
+      "step": 75
+    },
+    {
+      "completion_length": 77.38750076293945,
+      "epoch": 0.020307281229124917,
+      "grad_norm": 16.911163330078125,
+      "kl": 0.03070068359375,
+      "learning_rate": 9.796900053447354e-07,
+      "loss": 0.0012,
+      "reward": 0.39946897327899933,
+      "reward_std": 0.07368297874927521,
+      "rewards/iou_timestamp_reward": 0.39946897327899933,
+      "step": 76
+    },
+    {
+      "completion_length": 91.1500015258789,
+      "epoch": 0.020574482297929193,
+      "grad_norm": 23.92142105102539,
+      "kl": 0.0250244140625,
+      "learning_rate": 9.794227685729556e-07,
+      "loss": 0.001,
+      "reward": 0.6169492304325104,
+      "reward_std": 0.1484670452773571,
+      "rewards/iou_timestamp_reward": 0.6169492304325104,
+      "step": 77
+    },
+    {
+      "completion_length": 84.6624984741211,
+      "epoch": 0.020841683366733466,
+      "grad_norm": 7.9083685874938965,
+      "kl": 0.02764892578125,
+      "learning_rate": 9.791555318011759e-07,
+      "loss": 0.0011,
+      "reward": 0.3675663024187088,
+      "reward_std": 0.09683777764439583,
+      "rewards/iou_timestamp_reward": 0.3675663024187088,
+      "step": 78
+    },
+    {
+      "completion_length": 81.45000076293945,
+      "epoch": 0.021108884435537743,
+      "grad_norm": 31.15521812438965,
+      "kl": 0.0291748046875,
+      "learning_rate": 9.78888295029396e-07,
+      "loss": 0.0012,
+      "reward": 0.3303828537464142,
+      "reward_std": 0.08385182730853558,
+      "rewards/iou_timestamp_reward": 0.3303828537464142,
+      "step": 79
+    },
+    {
+      "completion_length": 79.38750076293945,
+      "epoch": 0.02137608550434202,
+      "grad_norm": 45.90896987915039,
+      "kl": 0.0291748046875,
+      "learning_rate": 9.786210582576162e-07,
+      "loss": 0.0012,
+      "reward": 0.5361809432506561,
+      "reward_std": 0.1461251862347126,
+      "rewards/iou_timestamp_reward": 0.5361809432506561,
+      "step": 80
+    },
+    {
+      "completion_length": 82.07500076293945,
+      "epoch": 0.021643286573146292,
+      "grad_norm": 305.8792419433594,
+      "kl": 0.02691650390625,
+      "learning_rate": 9.783538214858364e-07,
+      "loss": 0.0011,
+      "reward": 0.3591422736644745,
+      "reward_std": 0.06028897315263748,
+      "rewards/iou_timestamp_reward": 0.3591422736644745,
+      "step": 81
+    },
+    {
+      "completion_length": 85.0875015258789,
+      "epoch": 0.02191048764195057,
+      "grad_norm": 37.03975296020508,
+      "kl": 0.02716064453125,
+      "learning_rate": 9.780865847140567e-07,
+      "loss": 0.0011,
+      "reward": 0.5269813537597656,
+      "reward_std": 0.13692177832126617,
+      "rewards/iou_timestamp_reward": 0.5269813537597656,
+      "step": 82
+    },
+    {
+      "completion_length": 85.3499984741211,
+      "epoch": 0.02217768871075484,
+      "grad_norm": 44.27296447753906,
+      "kl": 0.0279541015625,
+      "learning_rate": 9.77819347942277e-07,
+      "loss": 0.0011,
+      "reward": 0.5033409297466278,
+      "reward_std": 0.14104880392551422,
+      "rewards/iou_timestamp_reward": 0.5033409297466278,
+      "step": 83
+    },
+    {
+      "completion_length": 82.22500228881836,
+      "epoch": 0.022444889779559118,
+      "grad_norm": 14.83884048461914,
+      "kl": 0.0341796875,
+      "learning_rate": 9.77552111170497e-07,
+      "loss": 0.0014,
+      "reward": 0.3850397616624832,
+      "reward_std": 0.07732919976115227,
+      "rewards/iou_timestamp_reward": 0.3850397616624832,
+      "step": 84
+    },
+    {
+      "completion_length": 81.88750076293945,
+      "epoch": 0.022712090848363394,
+      "grad_norm": 10.014225959777832,
+      "kl": 0.02880859375,
+      "learning_rate": 9.772848743987172e-07,
+      "loss": 0.0012,
+      "reward": 0.5256166011095047,
+      "reward_std": 0.0673129353672266,
+      "rewards/iou_timestamp_reward": 0.5256166011095047,
+      "step": 85
+    },
+    {
+      "completion_length": 84.23749923706055,
+      "epoch": 0.022979291917167667,
+      "grad_norm": 14.976663589477539,
+      "kl": 0.021484375,
+      "learning_rate": 9.770176376269374e-07,
+      "loss": 0.0009,
+      "reward": 0.4509493559598923,
+      "reward_std": 0.10696043819189072,
+      "rewards/iou_timestamp_reward": 0.4509493559598923,
+      "step": 86
+    },
+    {
+      "completion_length": 78.01250457763672,
+      "epoch": 0.023246492985971944,
+      "grad_norm": 15.414386749267578,
+      "kl": 0.02142333984375,
+      "learning_rate": 9.767504008551577e-07,
+      "loss": 0.0009,
+      "reward": 0.58517786860466,
+      "reward_std": 0.13582511618733406,
+      "rewards/iou_timestamp_reward": 0.58517786860466,
+      "step": 87
+    },
+    {
+      "completion_length": 87.87500381469727,
+      "epoch": 0.02351369405477622,
+      "grad_norm": 7.675116062164307,
+      "kl": 0.02691650390625,
+      "learning_rate": 9.76483164083378e-07,
+      "loss": 0.0011,
+      "reward": 0.4487306475639343,
+      "reward_std": 0.14682594686746597,
+      "rewards/iou_timestamp_reward": 0.4487306475639343,
+      "step": 88
+    },
+    {
+      "completion_length": 81.05000305175781,
+      "epoch": 0.023780895123580493,
+      "grad_norm": 13.316726684570312,
+      "kl": 0.02508544921875,
+      "learning_rate": 9.762159273115982e-07,
+      "loss": 0.001,
+      "reward": 0.707331657409668,
+      "reward_std": 0.11380118876695633,
+      "rewards/iou_timestamp_reward": 0.707331657409668,
+      "step": 89
+    },
+    {
+      "completion_length": 89.36250305175781,
+      "epoch": 0.02404809619238477,
+      "grad_norm": 14.177164077758789,
+      "kl": 0.02587890625,
+      "learning_rate": 9.759486905398182e-07,
+      "loss": 0.001,
+      "reward": 0.5567902773618698,
+      "reward_std": 0.11174913495779037,
+      "rewards/iou_timestamp_reward": 0.5567902773618698,
+      "step": 90
+    },
+    {
+      "completion_length": 94.0875015258789,
+      "epoch": 0.024315297261189046,
+      "grad_norm": 12.30399227142334,
+      "kl": 0.02972412109375,
+      "learning_rate": 9.756814537680384e-07,
+      "loss": 0.0012,
+      "reward": 0.503425657749176,
+      "reward_std": 0.10230003297328949,
+      "rewards/iou_timestamp_reward": 0.503425657749176,
+      "step": 91
+    },
+    {
+      "completion_length": 78.0875015258789,
+      "epoch": 0.02458249832999332,
+      "grad_norm": 24.13161277770996,
+      "kl": 0.0277099609375,
+      "learning_rate": 9.754142169962587e-07,
+      "loss": 0.0011,
+      "reward": 0.24431391805410385,
+      "reward_std": 0.07724586687982082,
+      "rewards/iou_timestamp_reward": 0.24431391805410385,
+      "step": 92
+    },
+    {
+      "completion_length": 85.22500228881836,
+      "epoch": 0.024849699398797595,
+      "grad_norm": 10.364203453063965,
+      "kl": 0.0269775390625,
+      "learning_rate": 9.75146980224479e-07,
+      "loss": 0.0011,
+      "reward": 0.5370314419269562,
+      "reward_std": 0.11360052973031998,
+      "rewards/iou_timestamp_reward": 0.5370314419269562,
+      "step": 93
+    },
+    {
+      "completion_length": 81.45000457763672,
+      "epoch": 0.02511690046760187,
+      "grad_norm": 12.295506477355957,
+      "kl": 0.0411376953125,
+      "learning_rate": 9.748797434526992e-07,
+      "loss": 0.0016,
+      "reward": 0.5579471290111542,
+      "reward_std": 0.09515488892793655,
+      "rewards/iou_timestamp_reward": 0.5579471290111542,
+      "step": 94
+    },
+    {
+      "completion_length": 80.38750076293945,
+      "epoch": 0.025384101536406144,
+      "grad_norm": 15.645418167114258,
+      "kl": 0.03857421875,
+      "learning_rate": 9.746125066809192e-07,
+      "loss": 0.0015,
+      "reward": 0.5186438858509064,
+      "reward_std": 0.11734926328063011,
+      "rewards/iou_timestamp_reward": 0.5186438858509064,
+      "step": 95
+    },
+    {
+      "completion_length": 81.93750381469727,
+      "epoch": 0.02565130260521042,
+      "grad_norm": 9.69706916809082,
+      "kl": 0.032958984375,
+      "learning_rate": 9.743452699091394e-07,
+      "loss": 0.0013,
+      "reward": 0.4298373907804489,
+      "reward_std": 0.12638958543539047,
+      "rewards/iou_timestamp_reward": 0.4298373907804489,
+      "step": 96
+    },
+    {
+      "completion_length": 86.4124984741211,
+      "epoch": 0.025918503674014697,
+      "grad_norm": 10.817190170288086,
+      "kl": 0.03369140625,
+      "learning_rate": 9.740780331373597e-07,
+      "loss": 0.0013,
+      "reward": 0.4257134795188904,
+      "reward_std": 0.0996755063533783,
+      "rewards/iou_timestamp_reward": 0.4257134795188904,
+      "step": 97
+    },
+    {
+      "completion_length": 87.60000228881836,
+      "epoch": 0.02618570474281897,
+      "grad_norm": 13.339404106140137,
+      "kl": 0.03131103515625,
+      "learning_rate": 9.7381079636558e-07,
+      "loss": 0.0012,
+      "reward": 0.466649129986763,
+      "reward_std": 0.12218038365244865,
+      "rewards/iou_timestamp_reward": 0.466649129986763,
+      "step": 98
+    },
+    {
+      "completion_length": 83.7125015258789,
+      "epoch": 0.026452905811623247,
+      "grad_norm": 6.138172626495361,
+      "kl": 0.0277099609375,
+      "learning_rate": 9.735435595938002e-07,
+      "loss": 0.0011,
+      "reward": 0.3628492206335068,
+      "reward_std": 0.08555204793810844,
+      "rewards/iou_timestamp_reward": 0.3628492206335068,
+      "step": 99
+    },
+    {
+      "completion_length": 83.5250015258789,
+      "epoch": 0.026720106880427523,
+      "grad_norm": 12.114334106445312,
+      "kl": 0.02606201171875,
+      "learning_rate": 9.732763228220202e-07,
+      "loss": 0.001,
+      "reward": 0.47904497385025024,
+      "reward_std": 0.11454221606254578,
+      "rewards/iou_timestamp_reward": 0.47904497385025024,
+      "step": 100
+    },
+    {
+      "completion_length": 89.67500305175781,
+      "epoch": 0.026987307949231796,
+      "grad_norm": 7.100372314453125,
+      "kl": 0.025146484375,
+      "learning_rate": 9.730090860502405e-07,
+      "loss": 0.001,
+      "reward": 0.5214036107063293,
+      "reward_std": 0.09529952518641949,
+      "rewards/iou_timestamp_reward": 0.5214036107063293,
+      "step": 101
+    },
+    {
+      "completion_length": 88.9375,
+      "epoch": 0.027254509018036072,
+      "grad_norm": 13.426268577575684,
+      "kl": 0.0247802734375,
+      "learning_rate": 9.727418492784607e-07,
+      "loss": 0.001,
+      "reward": 0.45033982396125793,
+      "reward_std": 0.11235591396689415,
+      "rewards/iou_timestamp_reward": 0.45033982396125793,
+      "step": 102
+    },
+    {
+      "completion_length": 93.35000228881836,
+      "epoch": 0.02752171008684035,
+      "grad_norm": 9.743562698364258,
+      "kl": 0.029052734375,
+      "learning_rate": 9.724746125066807e-07,
+      "loss": 0.0012,
+      "reward": 0.32814496755599976,
+      "reward_std": 0.10226564481854439,
+      "rewards/iou_timestamp_reward": 0.32814496755599976,
+      "step": 103
+    },
+    {
+      "completion_length": 97.7874984741211,
+      "epoch": 0.02778891115564462,
+      "grad_norm": 11.970060348510742,
+      "kl": 0.0242919921875,
+      "learning_rate": 9.72207375734901e-07,
+      "loss": 0.001,
+      "reward": 0.47338829934597015,
+      "reward_std": 0.12420981377363205,
+      "rewards/iou_timestamp_reward": 0.47338829934597015,
+      "step": 104
+    },
+    {
+      "completion_length": 89.0250015258789,
+      "epoch": 0.028056112224448898,
+      "grad_norm": 8.114420890808105,
+      "kl": 0.02874755859375,
+      "learning_rate": 9.719401389631212e-07,
+      "loss": 0.0012,
+      "reward": 0.38226931542158127,
+      "reward_std": 0.1174611747264862,
+      "rewards/iou_timestamp_reward": 0.38226931542158127,
+      "step": 105
+    },
+    {
+      "completion_length": 93.92499923706055,
+      "epoch": 0.028323313293253174,
+      "grad_norm": 17.39056968688965,
+      "kl": 0.03118896484375,
+      "learning_rate": 9.716729021913415e-07,
+      "loss": 0.0012,
+      "reward": 0.4228970557451248,
+      "reward_std": 0.13135263323783875,
+      "rewards/iou_timestamp_reward": 0.4228970557451248,
+      "step": 106
+    },
+    {
+      "completion_length": 102.91250228881836,
+      "epoch": 0.028590514362057447,
+      "grad_norm": 17.785310745239258,
+      "kl": 0.0283203125,
+      "learning_rate": 9.714056654195617e-07,
+      "loss": 0.0011,
+      "reward": 0.5794477760791779,
+      "reward_std": 0.10743552818894386,
+      "rewards/iou_timestamp_reward": 0.5794477760791779,
+      "step": 107
+    },
+    {
+      "completion_length": 99.43750381469727,
+      "epoch": 0.028857715430861724,
+      "grad_norm": 19.4038028717041,
+      "kl": 0.0311279296875,
+      "learning_rate": 9.711384286477817e-07,
+      "loss": 0.0012,
+      "reward": 0.45486655831336975,
+      "reward_std": 0.10233093053102493,
+      "rewards/iou_timestamp_reward": 0.45486655831336975,
+      "step": 108
+    },
+    {
+      "completion_length": 95.72500228881836,
+      "epoch": 0.029124916499666,
+      "grad_norm": 23.040130615234375,
+      "kl": 0.030517578125,
+      "learning_rate": 9.70871191876002e-07,
+      "loss": 0.0012,
+      "reward": 0.40837743878364563,
+      "reward_std": 0.12567105516791344,
+      "rewards/iou_timestamp_reward": 0.40837743878364563,
+      "step": 109
+    },
+    {
+      "completion_length": 96.85000228881836,
+      "epoch": 0.029392117568470273,
+      "grad_norm": 8.903193473815918,
+      "kl": 0.030029296875,
+      "learning_rate": 9.706039551042222e-07,
+      "loss": 0.0012,
+      "reward": 0.5052085667848587,
+      "reward_std": 0.13449953123927116,
+      "rewards/iou_timestamp_reward": 0.5052085667848587,
+      "step": 110
+    },
+    {
+      "completion_length": 91.8125,
+      "epoch": 0.02965931863727455,
+      "grad_norm": 20.629566192626953,
+      "kl": 0.02423095703125,
+      "learning_rate": 9.703367183324425e-07,
+      "loss": 0.001,
+      "reward": 0.323476105928421,
+      "reward_std": 0.07313787564635277,
+      "rewards/iou_timestamp_reward": 0.323476105928421,
+      "step": 111
+    },
+    {
+      "completion_length": 107.6500015258789,
+      "epoch": 0.029926519706078826,
+      "grad_norm": 10.181814193725586,
+      "kl": 0.0274658203125,
+      "learning_rate": 9.700694815606627e-07,
+      "loss": 0.0011,
+      "reward": 0.5477354377508163,
+      "reward_std": 0.11607206612825394,
+      "rewards/iou_timestamp_reward": 0.5477354377508163,
+      "step": 112
+    },
+    {
+      "completion_length": 94.04999923706055,
+      "epoch": 0.0301937207748831,
+      "grad_norm": 16.322965621948242,
+      "kl": 0.0316162109375,
+      "learning_rate": 9.69802244788883e-07,
+      "loss": 0.0013,
+      "reward": 0.45577606558799744,
+      "reward_std": 0.10110274329781532,
+      "rewards/iou_timestamp_reward": 0.45577606558799744,
+      "step": 113
+    },
+    {
+      "completion_length": 95.125,
+      "epoch": 0.030460921843687375,
+      "grad_norm": 20.937755584716797,
+      "kl": 0.0269775390625,
+      "learning_rate": 9.69535008017103e-07,
+      "loss": 0.0011,
+      "reward": 0.5386803448200226,
+      "reward_std": 0.09999115392565727,
+      "rewards/iou_timestamp_reward": 0.5386803448200226,
+      "step": 114
+    },
+    {
+      "completion_length": 92.45000076293945,
+      "epoch": 0.03072812291249165,
+      "grad_norm": 17.845489501953125,
+      "kl": 0.02886962890625,
+      "learning_rate": 9.692677712453233e-07,
+      "loss": 0.0012,
+      "reward": 0.44140908122062683,
+      "reward_std": 0.12080599367618561,
+      "rewards/iou_timestamp_reward": 0.44140908122062683,
+      "step": 115
+    },
+    {
+      "completion_length": 91.2125015258789,
+      "epoch": 0.030995323981295925,
+      "grad_norm": 11.29624080657959,
+      "kl": 0.04010009765625,
+      "learning_rate": 9.690005344735435e-07,
+      "loss": 0.0016,
+      "reward": 0.4579700082540512,
+      "reward_std": 0.1115545816719532,
+      "rewards/iou_timestamp_reward": 0.4579700082540512,
+      "step": 116
+    },
+    {
+      "completion_length": 91.4375,
+      "epoch": 0.0312625250501002,
+      "grad_norm": 9.191158294677734,
+      "kl": 0.026611328125,
+      "learning_rate": 9.687332977017637e-07,
+      "loss": 0.0011,
+      "reward": 0.5297419875860214,
+      "reward_std": 0.08494473621249199,
+      "rewards/iou_timestamp_reward": 0.5297419875860214,
+      "step": 117
+    },
+    {
+      "completion_length": 86.70000076293945,
+      "epoch": 0.03152972611890448,
+      "grad_norm": 155.39776611328125,
+      "kl": 0.0263671875,
+      "learning_rate": 9.68466060929984e-07,
+      "loss": 0.0011,
+      "reward": 0.3678671717643738,
+      "reward_std": 0.11950035020709038,
+      "rewards/iou_timestamp_reward": 0.3678671717643738,
+      "step": 118
+    },
+    {
+      "completion_length": 90.4625015258789,
+      "epoch": 0.031796927187708754,
+      "grad_norm": 18.508604049682617,
+      "kl": 0.02783203125,
+      "learning_rate": 9.68198824158204e-07,
+      "loss": 0.0011,
+      "reward": 0.4056996703147888,
+      "reward_std": 0.19096750020980835,
+      "rewards/iou_timestamp_reward": 0.4056996703147888,
+      "step": 119
+    },
+    {
+      "completion_length": 84.625,
+      "epoch": 0.03206412825651302,
+      "grad_norm": 11.088163375854492,
+      "kl": 0.02911376953125,
+      "learning_rate": 9.679315873864243e-07,
+      "loss": 0.0012,
+      "reward": 0.4425441175699234,
+      "reward_std": 0.09214091300964355,
+      "rewards/iou_timestamp_reward": 0.4425441175699234,
+      "step": 120
+    },
+    {
+      "completion_length": 89.11250305175781,
+      "epoch": 0.0323313293253173,
+      "grad_norm": 20.468366622924805,
+      "kl": 0.0357666015625,
+      "learning_rate": 9.676643506146445e-07,
+      "loss": 0.0014,
+      "reward": 0.4712321311235428,
+      "reward_std": 0.10969820618629456,
+      "rewards/iou_timestamp_reward": 0.4712321311235428,
+      "step": 121
+    },
+    {
+      "completion_length": 90.73750305175781,
+      "epoch": 0.032598530394121576,
+      "grad_norm": 20.332138061523438,
+      "kl": 0.03021240234375,
+      "learning_rate": 9.673971138428648e-07,
+      "loss": 0.0012,
+      "reward": 0.5420858561992645,
+      "reward_std": 0.14077245816588402,
+      "rewards/iou_timestamp_reward": 0.5420858561992645,
+      "step": 122
+    },
+    {
+      "completion_length": 80.5374984741211,
+      "epoch": 0.03286573146292585,
+      "grad_norm": 11.35522747039795,
+      "kl": 0.03619384765625,
+      "learning_rate": 9.67129877071085e-07,
+      "loss": 0.0014,
+      "reward": 0.4410727024078369,
+      "reward_std": 0.13154280930757523,
+      "rewards/iou_timestamp_reward": 0.4410727024078369,
+      "step": 123
+    },
+    {
+      "completion_length": 80.17499923706055,
+      "epoch": 0.03313293253173013,
+      "grad_norm": 15.309946060180664,
+      "kl": 0.04296875,
+      "learning_rate": 9.66862640299305e-07,
+      "loss": 0.0017,
+      "reward": 0.4710334688425064,
+      "reward_std": 0.08776786364614964,
+      "rewards/iou_timestamp_reward": 0.4710334688425064,
+      "step": 124
+    },
+    {
+      "completion_length": 81.87500381469727,
+      "epoch": 0.033400133600534405,
+      "grad_norm": 11.466004371643066,
+      "kl": 0.02783203125,
+      "learning_rate": 9.665954035275253e-07,
+      "loss": 0.0011,
+      "reward": 0.4691763147711754,
+      "reward_std": 0.08689278736710548,
+      "rewards/iou_timestamp_reward": 0.4691763147711754,
+      "step": 125
+    },
+    {
+      "completion_length": 82.82500076293945,
+      "epoch": 0.033667334669338675,
+      "grad_norm": 34.043914794921875,
+      "kl": 0.0379638671875,
+      "learning_rate": 9.663281667557455e-07,
+      "loss": 0.0015,
+      "reward": 0.44279107451438904,
+      "reward_std": 0.12102306261658669,
+      "rewards/iou_timestamp_reward": 0.44279107451438904,
+      "step": 126
+    },
+    {
+      "completion_length": 86.63750076293945,
+      "epoch": 0.03393453573814295,
+      "grad_norm": 193.87367248535156,
+      "kl": 0.0396728515625,
+      "learning_rate": 9.660609299839658e-07,
+      "loss": 0.0016,
+      "reward": 0.5007461756467819,
+      "reward_std": 0.10710695758461952,
+      "rewards/iou_timestamp_reward": 0.5007461756467819,
+      "step": 127
+    },
+    {
+      "completion_length": 88.38750076293945,
+      "epoch": 0.03420173680694723,
+      "grad_norm": 12.08212947845459,
+      "kl": 0.0350341796875,
+      "learning_rate": 9.65793693212186e-07,
+      "loss": 0.0014,
+      "reward": 0.4938567131757736,
+      "reward_std": 0.12255796045064926,
+      "rewards/iou_timestamp_reward": 0.4938567131757736,
+      "step": 128
+    },
+    {
+      "completion_length": 82.0625,
+      "epoch": 0.034468937875751504,
+      "grad_norm": 12.13739013671875,
+      "kl": 0.0396728515625,
+      "learning_rate": 9.655264564404063e-07,
+      "loss": 0.0016,
+      "reward": 0.4408050402998924,
+      "reward_std": 0.07024678587913513,
+      "rewards/iou_timestamp_reward": 0.4408050402998924,
+      "step": 129
+    },
+    {
+      "completion_length": 89.25000381469727,
+      "epoch": 0.03473613894455578,
+      "grad_norm": 9.668633460998535,
+      "kl": 0.0413818359375,
+      "learning_rate": 9.652592196686263e-07,
+      "loss": 0.0017,
+      "reward": 0.5121255666017532,
+      "reward_std": 0.08106988109648228,
+      "rewards/iou_timestamp_reward": 0.5121255666017532,
+      "step": 130
+    },
+    {
+      "completion_length": 83.66250228881836,
+      "epoch": 0.03500334001336005,
+      "grad_norm": 13.181618690490723,
+      "kl": 0.03662109375,
+      "learning_rate": 9.649919828968465e-07,
+      "loss": 0.0015,
+      "reward": 0.5275007039308548,
+      "reward_std": 0.07400373369455338,
+      "rewards/iou_timestamp_reward": 0.5275007039308548,
+      "step": 131
+    },
+    {
+      "completion_length": 87.2249984741211,
+      "epoch": 0.035270541082164326,
+      "grad_norm": 12.78144645690918,
+      "kl": 0.04296875,
+      "learning_rate": 9.647247461250668e-07,
+      "loss": 0.0017,
+      "reward": 0.5786421597003937,
+      "reward_std": 0.12599972262978554,
+      "rewards/iou_timestamp_reward": 0.5786421597003937,
+      "step": 132
+    },
+    {
+      "completion_length": 89.05000305175781,
+      "epoch": 0.0355377421509686,
+      "grad_norm": 14.639240264892578,
+      "kl": 0.0394287109375,
+      "learning_rate": 9.64457509353287e-07,
+      "loss": 0.0016,
+      "reward": 0.391816645860672,
+      "reward_std": 0.11170131340622902,
+      "rewards/iou_timestamp_reward": 0.391816645860672,
+      "step": 133
+    },
+    {
+      "completion_length": 92.57500076293945,
+      "epoch": 0.03580494321977288,
+      "grad_norm": 61.03050231933594,
+      "kl": 0.0511474609375,
+      "learning_rate": 9.641902725815073e-07,
+      "loss": 0.002,
+      "reward": 0.49499718844890594,
+      "reward_std": 0.11343395337462425,
+      "rewards/iou_timestamp_reward": 0.49499718844890594,
+      "step": 134
+    },
+    {
+      "completion_length": 84.36250305175781,
+      "epoch": 0.036072144288577156,
+      "grad_norm": 16.839923858642578,
+      "kl": 0.0423583984375,
+      "learning_rate": 9.639230358097273e-07,
+      "loss": 0.0017,
+      "reward": 0.3318554311990738,
+      "reward_std": 0.05935862101614475,
+      "rewards/iou_timestamp_reward": 0.3318554311990738,
+      "step": 135
+    },
+    {
+      "completion_length": 85.5875015258789,
+      "epoch": 0.03633934535738143,
+      "grad_norm": 14.992598533630371,
+      "kl": 0.037353515625,
+      "learning_rate": 9.636557990379475e-07,
+      "loss": 0.0015,
+      "reward": 0.4685795456171036,
+      "reward_std": 0.06687245890498161,
+      "rewards/iou_timestamp_reward": 0.4685795456171036,
+      "step": 136
+    },
+    {
+      "completion_length": 88.4000015258789,
+      "epoch": 0.0366065464261857,
+      "grad_norm": 17.69700813293457,
+      "kl": 0.038330078125,
+      "learning_rate": 9.633885622661678e-07,
+      "loss": 0.0015,
+      "reward": 0.6672376096248627,
+      "reward_std": 0.05268973857164383,
+      "rewards/iou_timestamp_reward": 0.6672376096248627,
+      "step": 137
+    },
+    {
+      "completion_length": 87.1500015258789,
+      "epoch": 0.03687374749498998,
+      "grad_norm": 12.710575103759766,
+      "kl": 0.0401611328125,
+      "learning_rate": 9.63121325494388e-07,
+      "loss": 0.0016,
+      "reward": 0.5832545459270477,
+      "reward_std": 0.10356196388602257,
+      "rewards/iou_timestamp_reward": 0.5832545459270477,
+      "step": 138
+    },
+    {
+      "completion_length": 89.20000076293945,
+      "epoch": 0.037140948563794254,
+      "grad_norm": 16.954652786254883,
+      "kl": 0.038818359375,
+      "learning_rate": 9.628540887226083e-07,
+      "loss": 0.0016,
+      "reward": 0.3957640081644058,
+      "reward_std": 0.09107673540711403,
+      "rewards/iou_timestamp_reward": 0.3957640081644058,
+      "step": 139
+    },
+    {
+      "completion_length": 89.10000228881836,
+      "epoch": 0.03740814963259853,
+      "grad_norm": 10.147904396057129,
+      "kl": 0.041748046875,
+      "learning_rate": 9.625868519508283e-07,
+      "loss": 0.0017,
+      "reward": 0.5282624512910843,
+      "reward_std": 0.10442003235220909,
+      "rewards/iou_timestamp_reward": 0.5282624512910843,
+      "step": 140
+    },
+    {
+      "completion_length": 84.7249984741211,
+      "epoch": 0.03767535070140281,
+      "grad_norm": 14.729109764099121,
+      "kl": 0.0543212890625,
+      "learning_rate": 9.623196151790486e-07,
+      "loss": 0.0022,
+      "reward": 0.4562620669603348,
+      "reward_std": 0.09092384111136198,
+      "rewards/iou_timestamp_reward": 0.4562620669603348,
+      "step": 141
+    },
+    {
+      "completion_length": 89.50000381469727,
+      "epoch": 0.037942551770207084,
+      "grad_norm": 17.36848258972168,
+      "kl": 0.038330078125,
+      "learning_rate": 9.620523784072688e-07,
+      "loss": 0.0015,
+      "reward": 0.5253326147794724,
+      "reward_std": 0.11369239538908005,
+      "rewards/iou_timestamp_reward": 0.5253326147794724,
+      "step": 142
+    },
+    {
+      "completion_length": 82.37500381469727,
+      "epoch": 0.03820975283901135,
+      "grad_norm": 10.699419975280762,
+      "kl": 0.0457763671875,
+      "learning_rate": 9.61785141635489e-07,
+      "loss": 0.0018,
+      "reward": 0.29642877727746964,
+      "reward_std": 0.052388858050107956,
+      "rewards/iou_timestamp_reward": 0.29642877727746964,
+      "step": 143
+    },
+    {
+      "completion_length": 91.28750228881836,
+      "epoch": 0.03847695390781563,
+      "grad_norm": 57.955970764160156,
+      "kl": 0.03057861328125,
+      "learning_rate": 9.615179048637093e-07,
+      "loss": 0.0012,
+      "reward": 0.49990953505039215,
+      "reward_std": 0.09969761222600937,
+      "rewards/iou_timestamp_reward": 0.49990953505039215,
+      "step": 144
+    },
+    {
+      "completion_length": 88.41250228881836,
+      "epoch": 0.038744154976619906,
+      "grad_norm": 23.626953125,
+      "kl": 0.0404052734375,
+      "learning_rate": 9.612506680919295e-07,
+      "loss": 0.0016,
+      "reward": 0.41132599115371704,
+      "reward_std": 0.07006280310451984,
+      "rewards/iou_timestamp_reward": 0.41132599115371704,
+      "step": 145
+    },
+    {
+      "completion_length": 90.97500228881836,
+      "epoch": 0.03901135604542418,
+      "grad_norm": 19.23369598388672,
+      "kl": 0.0394287109375,
+      "learning_rate": 9.609834313201496e-07,
+      "loss": 0.0016,
+      "reward": 0.4373021870851517,
+      "reward_std": 0.12405096367001534,
+      "rewards/iou_timestamp_reward": 0.4373021870851517,
+      "step": 146
+    },
+    {
+      "completion_length": 84.13750076293945,
+      "epoch": 0.03927855711422846,
+      "grad_norm": 9.589532852172852,
+      "kl": 0.039794921875,
+      "learning_rate": 9.607161945483698e-07,
+      "loss": 0.0016,
+      "reward": 0.409616619348526,
+      "reward_std": 0.11418266966938972,
+      "rewards/iou_timestamp_reward": 0.409616619348526,
+      "step": 147
+    },
+    {
+      "completion_length": 78.5875015258789,
+      "epoch": 0.039545758183032735,
+      "grad_norm": 14.857934951782227,
+      "kl": 0.0352783203125,
+      "learning_rate": 9.6044895777659e-07,
+      "loss": 0.0014,
+      "reward": 0.5202317535877228,
+      "reward_std": 0.11668656021356583,
+      "rewards/iou_timestamp_reward": 0.5202317535877228,
+      "step": 148
+    },
+    {
+      "completion_length": 82.8125,
+      "epoch": 0.039812959251837005,
+      "grad_norm": 10.997519493103027,
+      "kl": 0.03094482421875,
+      "learning_rate": 9.601817210048103e-07,
+      "loss": 0.0012,
+      "reward": 0.42350925505161285,
+      "reward_std": 0.12441174313426018,
+      "rewards/iou_timestamp_reward": 0.42350925505161285,
+      "step": 149
+    },
+    {
+      "completion_length": 79.18750381469727,
+      "epoch": 0.04008016032064128,
+      "grad_norm": 11.452859878540039,
+      "kl": 0.0384521484375,
+      "learning_rate": 9.599144842330306e-07,
+      "loss": 0.0015,
+      "reward": 0.42812803387641907,
+      "reward_std": 0.07005837932229042,
+      "rewards/iou_timestamp_reward": 0.42812803387641907,
+      "step": 150
+    },
+    {
+      "completion_length": 85.4375,
+      "epoch": 0.04034736138944556,
+      "grad_norm": 21.108501434326172,
+      "kl": 0.0325927734375,
+      "learning_rate": 9.596472474612506e-07,
+      "loss": 0.0013,
+      "reward": 0.3734997771680355,
+      "reward_std": 0.10468435287475586,
+      "rewards/iou_timestamp_reward": 0.3734997771680355,
+      "step": 151
+    },
+    {
+      "completion_length": 83.7750015258789,
+      "epoch": 0.040614562458249834,
+      "grad_norm": 17.950271606445312,
+      "kl": 0.045654296875,
+      "learning_rate": 9.593800106894708e-07,
+      "loss": 0.0018,
+      "reward": 0.40336285531520844,
+      "reward_std": 0.08605783060193062,
+      "rewards/iou_timestamp_reward": 0.40336285531520844,
+      "step": 152
+    },
+    {
+      "completion_length": 80.5875015258789,
+      "epoch": 0.04088176352705411,
+      "grad_norm": 15.713683128356934,
+      "kl": 0.03759765625,
+      "learning_rate": 9.59112773917691e-07,
+      "loss": 0.0015,
+      "reward": 0.4455360919237137,
+      "reward_std": 0.10942557081580162,
+      "rewards/iou_timestamp_reward": 0.4455360919237137,
+      "step": 153
+    },
+    {
+      "completion_length": 93.82500076293945,
+      "epoch": 0.04114896459585839,
+      "grad_norm": 130.20535278320312,
+      "kl": 0.048583984375,
+      "learning_rate": 9.588455371459111e-07,
+      "loss": 0.0019,
+      "reward": 0.47216907143592834,
+      "reward_std": 0.12527307868003845,
+      "rewards/iou_timestamp_reward": 0.47216907143592834,
+      "step": 154
+    },
+    {
+      "completion_length": 77.4375,
+      "epoch": 0.041416165664662656,
+      "grad_norm": 33.787925720214844,
+      "kl": 0.044189453125,
+      "learning_rate": 9.585783003741314e-07,
+      "loss": 0.0018,
+      "reward": 0.5094977915287018,
+      "reward_std": 0.11581655591726303,
+      "rewards/iou_timestamp_reward": 0.5094977915287018,
+      "step": 155
+    },
+    {
+      "completion_length": 81.1500015258789,
+      "epoch": 0.04168336673346693,
+      "grad_norm": 14.96328353881836,
+      "kl": 0.0411376953125,
+      "learning_rate": 9.583110636023516e-07,
+      "loss": 0.0016,
+      "reward": 0.39587974548339844,
+      "reward_std": 0.11371102556586266,
+      "rewards/iou_timestamp_reward": 0.39587974548339844,
+      "step": 156
+    },
+    {
+      "completion_length": 90.51250076293945,
+      "epoch": 0.04195056780227121,
+      "grad_norm": 12.745133399963379,
+      "kl": 0.03662109375,
+      "learning_rate": 9.580438268305718e-07,
+      "loss": 0.0015,
+      "reward": 0.4054044932126999,
+      "reward_std": 0.06693214736878872,
+      "rewards/iou_timestamp_reward": 0.4054044932126999,
+      "step": 157
+    },
+    {
+      "completion_length": 85.22500228881836,
+      "epoch": 0.042217768871075485,
+      "grad_norm": 15.326485633850098,
+      "kl": 0.04296875,
+      "learning_rate": 9.57776590058792e-07,
+      "loss": 0.0017,
+      "reward": 0.517610028386116,
+      "reward_std": 0.08778567984700203,
+      "rewards/iou_timestamp_reward": 0.517610028386116,
+      "step": 158
+    },
+    {
+      "completion_length": 85.10000228881836,
+      "epoch": 0.04248496993987976,
+      "grad_norm": 15.57522964477539,
+      "kl": 0.0545654296875,
+      "learning_rate": 9.575093532870121e-07,
+      "loss": 0.0022,
+      "reward": 0.47441329061985016,
+      "reward_std": 0.11022436246275902,
+      "rewards/iou_timestamp_reward": 0.47441329061985016,
+      "step": 159
+    },
+    {
+      "completion_length": 80.95000076293945,
+      "epoch": 0.04275217100868404,
+      "grad_norm": 14.413505554199219,
+      "kl": 0.0439453125,
+      "learning_rate": 9.572421165152324e-07,
+      "loss": 0.0018,
+      "reward": 0.38343895971775055,
+      "reward_std": 0.1145082600414753,
+      "rewards/iou_timestamp_reward": 0.38343895971775055,
+      "step": 160
+    },
+    {
+      "completion_length": 79.8375015258789,
+      "epoch": 0.04301937207748831,
+      "grad_norm": 12.668608665466309,
+      "kl": 0.037109375,
+      "learning_rate": 9.569748797434526e-07,
+      "loss": 0.0015,
+      "reward": 0.5705928802490234,
+      "reward_std": 0.10167951509356499,
+      "rewards/iou_timestamp_reward": 0.5705928802490234,
+      "step": 161
+    },
+    {
+      "completion_length": 77.04999923706055,
+      "epoch": 0.043286573146292584,
+      "grad_norm": 17.551441192626953,
+      "kl": 0.0455322265625,
+      "learning_rate": 9.567076429716729e-07,
+      "loss": 0.0018,
+      "reward": 0.5757200121879578,
+      "reward_std": 0.12645361945033073,
+      "rewards/iou_timestamp_reward": 0.5757200121879578,
+      "step": 162
+    },
+    {
+      "completion_length": 76.67500305175781,
+      "epoch": 0.04355377421509686,
+      "grad_norm": 12.472587585449219,
+      "kl": 0.043701171875,
+      "learning_rate": 9.56440406199893e-07,
+      "loss": 0.0017,
+      "reward": 0.3930691331624985,
+      "reward_std": 0.047056157141923904,
+      "rewards/iou_timestamp_reward": 0.3930691331624985,
+      "step": 163
+    },
+    {
+      "completion_length": 82.97500228881836,
+      "epoch": 0.04382097528390114,
+      "grad_norm": 41.681278228759766,
+      "kl": 0.050048828125,
+      "learning_rate": 9.561731694281131e-07,
+      "loss": 0.002,
+      "reward": 0.37715357542037964,
+      "reward_std": 0.14025865495204926,
+      "rewards/iou_timestamp_reward": 0.37715357542037964,
+      "step": 164
+    },
+    {
+      "completion_length": 81.3499984741211,
+      "epoch": 0.04408817635270541,
+      "grad_norm": 13.00397777557373,
+      "kl": 0.0457763671875,
+      "learning_rate": 9.559059326563334e-07,
+      "loss": 0.0018,
+      "reward": 0.4659877270460129,
+      "reward_std": 0.07307687401771545,
+      "rewards/iou_timestamp_reward": 0.4659877270460129,
+      "step": 165
+    },
+    {
+      "completion_length": 79.07500076293945,
+      "epoch": 0.04435537742150968,
+      "grad_norm": 42.21580505371094,
+      "kl": 0.039794921875,
+      "learning_rate": 9.556386958845536e-07,
+      "loss": 0.0016,
+      "reward": 0.45951294898986816,
+      "reward_std": 0.11066334694623947,
+      "rewards/iou_timestamp_reward": 0.45951294898986816,
+      "step": 166
+    },
+    {
+      "completion_length": 82.25,
+      "epoch": 0.04462257849031396,
+      "grad_norm": 16.60811996459961,
+      "kl": 0.0499267578125,
+      "learning_rate": 9.553714591127739e-07,
+      "loss": 0.002,
+      "reward": 0.5194784104824066,
+      "reward_std": 0.09066024795174599,
+      "rewards/iou_timestamp_reward": 0.5194784104824066,
+      "step": 167
+    },
+    {
+      "completion_length": 79.26250076293945,
+      "epoch": 0.044889779559118236,
+      "grad_norm": 43.955142974853516,
+      "kl": 0.0406494140625,
+      "learning_rate": 9.551042223409941e-07,
+      "loss": 0.0016,
+      "reward": 0.40560464560985565,
+      "reward_std": 0.07998048886656761,
+      "rewards/iou_timestamp_reward": 0.40560464560985565,
+      "step": 168
+    },
+    {
+      "completion_length": 78.88750076293945,
+      "epoch": 0.04515698062792251,
+      "grad_norm": 10.694327354431152,
+      "kl": 0.0501708984375,
+      "learning_rate": 9.548369855692144e-07,
+      "loss": 0.002,
+      "reward": 0.5325513929128647,
+      "reward_std": 0.05170968174934387,
+      "rewards/iou_timestamp_reward": 0.5325513929128647,
+      "step": 169
+    },
+    {
+      "completion_length": 75.86250305175781,
+      "epoch": 0.04542418169672679,
+      "grad_norm": 7.294441223144531,
+      "kl": 0.03662109375,
+      "learning_rate": 9.545697487974344e-07,
+      "loss": 0.0015,
+      "reward": 0.46570850908756256,
+      "reward_std": 0.07274461165070534,
+      "rewards/iou_timestamp_reward": 0.46570850908756256,
+      "step": 170
+    },
+    {
+      "completion_length": 75.23750305175781,
+      "epoch": 0.045691382765531065,
+      "grad_norm": 13.37921142578125,
+      "kl": 0.0406494140625,
+      "learning_rate": 9.543025120256546e-07,
+      "loss": 0.0016,
+      "reward": 0.4209117740392685,
+      "reward_std": 0.10308519750833511,
+      "rewards/iou_timestamp_reward": 0.4209117740392685,
+      "step": 171
+    },
+    {
+      "completion_length": 75.0250015258789,
+      "epoch": 0.045958583834335334,
+      "grad_norm": 19.120006561279297,
+      "kl": 0.0513916015625,
+      "learning_rate": 9.540352752538749e-07,
+      "loss": 0.0021,
+      "reward": 0.5029748231172562,
+      "reward_std": 0.06725198961794376,
+      "rewards/iou_timestamp_reward": 0.5029748231172562,
+      "step": 172
+    },
+    {
+      "completion_length": 76.5625,
+      "epoch": 0.04622578490313961,
+      "grad_norm": 11.126622200012207,
+      "kl": 0.0367431640625,
+      "learning_rate": 9.537680384820951e-07,
+      "loss": 0.0015,
+      "reward": 0.5426850020885468,
+      "reward_std": 0.07354238256812096,
+      "rewards/iou_timestamp_reward": 0.5426850020885468,
+      "step": 173
+    },
+    {
+      "completion_length": 81.98749923706055,
+      "epoch": 0.04649298597194389,
+      "grad_norm": 15.441088676452637,
+      "kl": 0.04541015625,
+      "learning_rate": 9.535008017103154e-07,
+      "loss": 0.0018,
+      "reward": 0.366536945104599,
+      "reward_std": 0.10370578989386559,
+      "rewards/iou_timestamp_reward": 0.366536945104599,
+      "step": 174
+    },
+    {
+      "completion_length": 76.9124984741211,
+      "epoch": 0.04676018704074816,
+      "grad_norm": 18.660261154174805,
+      "kl": 0.0594482421875,
+      "learning_rate": 9.532335649385355e-07,
+      "loss": 0.0024,
+      "reward": 0.534787967801094,
+      "reward_std": 0.10525619983673096,
+      "rewards/iou_timestamp_reward": 0.534787967801094,
+      "step": 175
+    },
+    {
+      "completion_length": 77.11249923706055,
+      "epoch": 0.04702738810955244,
+      "grad_norm": 9.616857528686523,
+      "kl": 0.0528564453125,
+      "learning_rate": 9.529663281667558e-07,
+      "loss": 0.0021,
+      "reward": 0.46122419834136963,
+      "reward_std": 0.11353456228971481,
+      "rewards/iou_timestamp_reward": 0.46122419834136963,
+      "step": 176
+    },
+    {
+      "completion_length": 76.57500457763672,
+      "epoch": 0.047294589178356716,
+      "grad_norm": 11.133618354797363,
+      "kl": 0.0484619140625,
+      "learning_rate": 9.52699091394976e-07,
+      "loss": 0.0019,
+      "reward": 0.4199577569961548,
+      "reward_std": 0.09651762992143631,
+      "rewards/iou_timestamp_reward": 0.4199577569961548,
+      "step": 177
+    },
+    {
+      "completion_length": 79.05000305175781,
+      "epoch": 0.047561790247160986,
+      "grad_norm": 12.524818420410156,
+      "kl": 0.049560546875,
+      "learning_rate": 9.52431854623196e-07,
+      "loss": 0.002,
+      "reward": 0.5286326110363007,
+      "reward_std": 0.12330913916230202,
+      "rewards/iou_timestamp_reward": 0.5286326110363007,
+      "step": 178
+    },
+    {
+      "completion_length": 83.50000381469727,
+      "epoch": 0.04782899131596526,
+      "grad_norm": 12.21944808959961,
+      "kl": 0.0408935546875,
+      "learning_rate": 9.521646178514163e-07,
+      "loss": 0.0016,
+      "reward": 0.47760356962680817,
+      "reward_std": 0.08412851020693779,
+      "rewards/iou_timestamp_reward": 0.47760356962680817,
+      "step": 179
+    },
+    {
+      "completion_length": 74.3375015258789,
+      "epoch": 0.04809619238476954,
+      "grad_norm": 15.917021751403809,
+      "kl": 0.0504150390625,
+      "learning_rate": 9.518973810796364e-07,
+      "loss": 0.002,
+      "reward": 0.4619504362344742,
+      "reward_std": 0.06471715308725834,
+      "rewards/iou_timestamp_reward": 0.4619504362344742,
+      "step": 180
+    },
+    {
+      "completion_length": 85.70000076293945,
+      "epoch": 0.048363393453573815,
+      "grad_norm": 17.021228790283203,
+      "kl": 0.0552978515625,
+      "learning_rate": 9.516301443078567e-07,
+      "loss": 0.0022,
+      "reward": 0.3753501623868942,
+      "reward_std": 0.08277348056435585,
+      "rewards/iou_timestamp_reward": 0.3753501623868942,
+      "step": 181
+    },
+    {
+      "completion_length": 86.86249923706055,
+      "epoch": 0.04863059452237809,
+      "grad_norm": 49.27531051635742,
+      "kl": 0.073974609375,
+      "learning_rate": 9.513629075360769e-07,
+      "loss": 0.003,
+      "reward": 0.644359678030014,
+      "reward_std": 0.10561589896678925,
+      "rewards/iou_timestamp_reward": 0.644359678030014,
+      "step": 182
+    },
+    {
+      "completion_length": 79.38750076293945,
+      "epoch": 0.04889779559118237,
+      "grad_norm": 94.31536102294922,
+      "kl": 0.0479736328125,
+      "learning_rate": 9.51095670764297e-07,
+      "loss": 0.0019,
+      "reward": 0.6270514130592346,
+      "reward_std": 0.09841878339648247,
+      "rewards/iou_timestamp_reward": 0.6270514130592346,
+      "step": 183
+    },
+    {
+      "completion_length": 76.35000228881836,
+      "epoch": 0.04916499665998664,
+      "grad_norm": 19.030534744262695,
+      "kl": 0.059814453125,
+      "learning_rate": 9.508284339925173e-07,
+      "loss": 0.0024,
+      "reward": 0.5081672370433807,
+      "reward_std": 0.10591346770524979,
+      "rewards/iou_timestamp_reward": 0.5081672370433807,
+      "step": 184
+    },
+    {
+      "completion_length": 76.4124984741211,
+      "epoch": 0.049432197728790914,
+      "grad_norm": 18.321247100830078,
+      "kl": 0.046142578125,
+      "learning_rate": 9.505611972207375e-07,
+      "loss": 0.0018,
+      "reward": 0.46649011969566345,
+      "reward_std": 0.1195777915418148,
+      "rewards/iou_timestamp_reward": 0.46649011969566345,
+      "step": 185
+    },
+    {
+      "completion_length": 84.7125015258789,
+      "epoch": 0.04969939879759519,
+      "grad_norm": 13.643217086791992,
+      "kl": 0.047607421875,
+      "learning_rate": 9.502939604489577e-07,
+      "loss": 0.0019,
+      "reward": 0.5010816007852554,
+      "reward_std": 0.07837207242846489,
+      "rewards/iou_timestamp_reward": 0.5010816007852554,
+      "step": 186
+    },
+    {
+      "completion_length": 94.1500015258789,
+      "epoch": 0.049966599866399466,
+      "grad_norm": 18.298660278320312,
+      "kl": 0.049560546875,
+      "learning_rate": 9.500267236771779e-07,
+      "loss": 0.002,
+      "reward": 0.44392018020153046,
+      "reward_std": 0.07208472117781639,
+      "rewards/iou_timestamp_reward": 0.44392018020153046,
+      "step": 187
+    },
+    {
+      "completion_length": 83.9000015258789,
+      "epoch": 0.05023380093520374,
+      "grad_norm": 43.245094299316406,
+      "kl": 0.064208984375,
+      "learning_rate": 9.497594869053981e-07,
+      "loss": 0.0026,
+      "reward": 0.4399380534887314,
+      "reward_std": 0.07741782441735268,
+      "rewards/iou_timestamp_reward": 0.4399380534887314,
+      "step": 188
+    },
+    {
+      "completion_length": 90.23750305175781,
+      "epoch": 0.05050100200400802,
+      "grad_norm": 7.059505462646484,
+      "kl": 0.0338134765625,
+      "learning_rate": 9.494922501336183e-07,
+      "loss": 0.0014,
+      "reward": 0.4319871515035629,
+      "reward_std": 0.030242225155234337,
+      "rewards/iou_timestamp_reward": 0.4319871515035629,
+      "step": 189
+    },
+    {
+      "completion_length": 83.70000076293945,
+      "epoch": 0.05076820307281229,
+      "grad_norm": 17.03489875793457,
+      "kl": 0.0411376953125,
+      "learning_rate": 9.492250133618386e-07,
+      "loss": 0.0016,
+      "reward": 0.5623960196971893,
+      "reward_std": 0.10269278660416603,
+      "rewards/iou_timestamp_reward": 0.5623960196971893,
+      "step": 190
+    },
+    {
+      "completion_length": 90.26250076293945,
+      "epoch": 0.051035404141616565,
+      "grad_norm": 8.389057159423828,
+      "kl": 0.044677734375,
+      "learning_rate": 9.489577765900587e-07,
+      "loss": 0.0018,
+      "reward": 0.462777778506279,
+      "reward_std": 0.08947985991835594,
+      "rewards/iou_timestamp_reward": 0.462777778506279,
+      "step": 191
+    },
+    {
+      "completion_length": 88.85000228881836,
+      "epoch": 0.05130260521042084,
+      "grad_norm": 8.912788391113281,
+      "kl": 0.05419921875,
+      "learning_rate": 9.486905398182789e-07,
+      "loss": 0.0022,
+      "reward": 0.3437252789735794,
+      "reward_std": 0.1144870612770319,
+      "rewards/iou_timestamp_reward": 0.3437252789735794,
+      "step": 192
+    },
+    {
+      "completion_length": 80.23750305175781,
+      "epoch": 0.05156980627922512,
+      "grad_norm": 16.888032913208008,
+      "kl": 0.0465087890625,
+      "learning_rate": 9.484233030464992e-07,
+      "loss": 0.0019,
+      "reward": 0.32809334993362427,
+      "reward_std": 0.08787024766206741,
+      "rewards/iou_timestamp_reward": 0.32809334993362427,
+      "step": 193
+    },
+    {
+      "completion_length": 82.57500076293945,
+      "epoch": 0.051837007348029394,
+      "grad_norm": 122.45707702636719,
+      "kl": 0.05322265625,
+      "learning_rate": 9.481560662747193e-07,
+      "loss": 0.0021,
+      "reward": 0.39617256820201874,
+      "reward_std": 0.10278888791799545,
+      "rewards/iou_timestamp_reward": 0.39617256820201874,
+      "step": 194
+    },
+    {
+      "completion_length": 76.86250305175781,
+      "epoch": 0.052104208416833664,
+      "grad_norm": 21.19937515258789,
+      "kl": 0.0482177734375,
+      "learning_rate": 9.478888295029396e-07,
+      "loss": 0.0019,
+      "reward": 0.4638468474149704,
+      "reward_std": 0.1405925117433071,
+      "rewards/iou_timestamp_reward": 0.4638468474149704,
+      "step": 195
+    },
+    {
+      "completion_length": 83.17499923706055,
+      "epoch": 0.05237140948563794,
+      "grad_norm": 13.251508712768555,
+      "kl": 0.0523681640625,
+      "learning_rate": 9.476215927311597e-07,
+      "loss": 0.0021,
+      "reward": 0.485527366399765,
+      "reward_std": 0.09180990979075432,
+      "rewards/iou_timestamp_reward": 0.485527366399765,
+      "step": 196
+    },
+    {
+      "completion_length": 85.42499923706055,
+      "epoch": 0.05263861055444222,
+      "grad_norm": 9.114768028259277,
+      "kl": 0.042724609375,
+      "learning_rate": 9.4735435595938e-07,
+      "loss": 0.0017,
+      "reward": 0.4791525900363922,
+      "reward_std": 0.09207051619887352,
+      "rewards/iou_timestamp_reward": 0.4791525900363922,
+      "step": 197
+    },
+    {
+      "completion_length": 87.76250457763672,
+      "epoch": 0.05290581162324649,
+      "grad_norm": 11.511754035949707,
+      "kl": 0.05859375,
+      "learning_rate": 9.470871191876002e-07,
+      "loss": 0.0023,
+      "reward": 0.3516260161995888,
+      "reward_std": 0.12914332374930382,
+      "rewards/iou_timestamp_reward": 0.3516260161995888,
+      "step": 198
+    },
+    {
+      "completion_length": 77.92500305175781,
+      "epoch": 0.05317301269205077,
+      "grad_norm": 11.806519508361816,
+      "kl": 0.0458984375,
+      "learning_rate": 9.468198824158203e-07,
+      "loss": 0.0018,
+      "reward": 0.4304570257663727,
+      "reward_std": 0.04932829737663269,
+      "rewards/iou_timestamp_reward": 0.4304570257663727,
+      "step": 199
+    },
+    {
+      "completion_length": 81.9749984741211,
+      "epoch": 0.053440213760855046,
+      "grad_norm": 16.832813262939453,
+      "kl": 0.0489501953125,
+      "learning_rate": 9.465526456440406e-07,
+      "loss": 0.002,
+      "reward": 0.4915754646062851,
+      "reward_std": 0.13631991669535637,
+      "rewards/iou_timestamp_reward": 0.4915754646062851,
+      "step": 200
+    },
+    {
+      "completion_length": 85.57500457763672,
+      "epoch": 0.053707414829659315,
+      "grad_norm": 8.950471878051758,
+      "kl": 0.0538330078125,
+      "learning_rate": 9.462854088722608e-07,
+      "loss": 0.0022,
+      "reward": 0.32636383175849915,
+      "reward_std": 0.0875160414725542,
+      "rewards/iou_timestamp_reward": 0.32636383175849915,
+      "step": 201
+    },
+    {
+      "completion_length": 88.1875,
+      "epoch": 0.05397461589846359,
+      "grad_norm": 8.32710075378418,
+      "kl": 0.0411376953125,
+      "learning_rate": 9.46018172100481e-07,
+      "loss": 0.0016,
+      "reward": 0.3660304397344589,
+      "reward_std": 0.12842414155602455,
+      "rewards/iou_timestamp_reward": 0.3660304397344589,
+      "step": 202
+    },
+    {
+      "completion_length": 83.18750381469727,
+      "epoch": 0.05424181696726787,
+      "grad_norm": 34.32887649536133,
+      "kl": 0.0579833984375,
+      "learning_rate": 9.457509353287012e-07,
+      "loss": 0.0023,
+      "reward": 0.4675799459218979,
+      "reward_std": 0.11138150840997696,
+      "rewards/iou_timestamp_reward": 0.4675799459218979,
+      "step": 203
+    },
+    {
+      "completion_length": 78.57500076293945,
+      "epoch": 0.054509018036072145,
+      "grad_norm": 11.615337371826172,
+      "kl": 0.0482177734375,
+      "learning_rate": 9.454836985569213e-07,
+      "loss": 0.0019,
+      "reward": 0.20401198416948318,
+      "reward_std": 0.06908228248357773,
+      "rewards/iou_timestamp_reward": 0.20401198416948318,
+      "step": 204
+    },
+    {
+      "completion_length": 80.95000076293945,
+      "epoch": 0.05477621910487642,
+      "grad_norm": 15.079862594604492,
+      "kl": 0.0557861328125,
+      "learning_rate": 9.452164617851416e-07,
+      "loss": 0.0022,
+      "reward": 0.542988657951355,
+      "reward_std": 0.1345556452870369,
+      "rewards/iou_timestamp_reward": 0.542988657951355,
+      "step": 205
+    },
+    {
+      "completion_length": 71.92500305175781,
+      "epoch": 0.0550434201736807,
+      "grad_norm": 10.543124198913574,
+      "kl": 0.0556640625,
+      "learning_rate": 9.449492250133618e-07,
+      "loss": 0.0022,
+      "reward": 0.2738891839981079,
+      "reward_std": 0.10082607716321945,
+      "rewards/iou_timestamp_reward": 0.2738891839981079,
+      "step": 206
+    },
+    {
+      "completion_length": 75.97500228881836,
+      "epoch": 0.05531062124248497,
+      "grad_norm": 11.855889320373535,
+      "kl": 0.0440673828125,
+      "learning_rate": 9.44681988241582e-07,
+      "loss": 0.0018,
+      "reward": 0.40028373897075653,
+      "reward_std": 0.07233771868050098,
+      "rewards/iou_timestamp_reward": 0.40028373897075653,
+      "step": 207
+    },
+    {
+      "completion_length": 75.36250305175781,
+      "epoch": 0.05557782231128924,
+      "grad_norm": 11.384106636047363,
+      "kl": 0.0484619140625,
+      "learning_rate": 9.444147514698022e-07,
+      "loss": 0.0019,
+      "reward": 0.518831193447113,
+      "reward_std": 0.04942659102380276,
+      "rewards/iou_timestamp_reward": 0.518831193447113,
+      "step": 208
+    },
+    {
+      "completion_length": 83.93750381469727,
+      "epoch": 0.05584502338009352,
+      "grad_norm": 28.444765090942383,
+      "kl": 0.0517578125,
+      "learning_rate": 9.441475146980225e-07,
+      "loss": 0.0021,
+      "reward": 0.591288223862648,
+      "reward_std": 0.13365697488188744,
+      "rewards/iou_timestamp_reward": 0.591288223862648,
+      "step": 209
+    },
+    {
+      "completion_length": 77.67499923706055,
+      "epoch": 0.056112224448897796,
+      "grad_norm": 47.66985321044922,
+      "kl": 0.060791015625,
+      "learning_rate": 9.438802779262426e-07,
+      "loss": 0.0024,
+      "reward": 0.4540657252073288,
+      "reward_std": 0.11418235674500465,
+      "rewards/iou_timestamp_reward": 0.4540657252073288,
+      "step": 210
+    },
+    {
+      "completion_length": 85.98750305175781,
+      "epoch": 0.05637942551770207,
+      "grad_norm": 17.33128547668457,
+      "kl": 0.0565185546875,
+      "learning_rate": 9.436130411544628e-07,
+      "loss": 0.0023,
+      "reward": 0.45703037083148956,
+      "reward_std": 0.12524070590734482,
+      "rewards/iou_timestamp_reward": 0.45703037083148956,
+      "step": 211
+    },
+    {
+      "completion_length": 77.5,
+      "epoch": 0.05664662658650635,
+      "grad_norm": 9.057528495788574,
+      "kl": 0.05810546875,
+      "learning_rate": 9.43345804382683e-07,
+      "loss": 0.0023,
+      "reward": 0.4245501309633255,
+      "reward_std": 0.07445455808192492,
+      "rewards/iou_timestamp_reward": 0.4245501309633255,
+      "step": 212
+    },
+    {
+      "completion_length": 89.73750305175781,
+      "epoch": 0.05691382765531062,
+      "grad_norm": 9.169981956481934,
+      "kl": 0.08251953125,
+      "learning_rate": 9.430785676109032e-07,
+      "loss": 0.0033,
+      "reward": 0.5322041213512421,
+      "reward_std": 0.09248698502779007,
+      "rewards/iou_timestamp_reward": 0.5322041213512421,
+      "step": 213
+    },
+    {
+      "completion_length": 82.36250305175781,
+      "epoch": 0.057181028724114895,
+      "grad_norm": 13.932048797607422,
+      "kl": 0.069091796875,
+      "learning_rate": 9.428113308391235e-07,
+      "loss": 0.0028,
+      "reward": 0.43773867189884186,
+      "reward_std": 0.08092993497848511,
+      "rewards/iou_timestamp_reward": 0.43773867189884186,
+      "step": 214
+    },
+    {
+      "completion_length": 83.1624984741211,
+      "epoch": 0.05744822979291917,
+      "grad_norm": 19.159921646118164,
+      "kl": 0.0606689453125,
+      "learning_rate": 9.425440940673436e-07,
+      "loss": 0.0024,
+      "reward": 0.5727377682924271,
+      "reward_std": 0.07809071242809296,
+      "rewards/iou_timestamp_reward": 0.5727377682924271,
+      "step": 215
+    },
+    {
+      "completion_length": 75.51250076293945,
+      "epoch": 0.05771543086172345,
+      "grad_norm": 7.645944118499756,
+      "kl": 0.0511474609375,
+      "learning_rate": 9.422768572955639e-07,
+      "loss": 0.002,
+      "reward": 0.5318755507469177,
+      "reward_std": 0.053884051740169525,
+      "rewards/iou_timestamp_reward": 0.5318755507469177,
+      "step": 216
+    },
+    {
+      "completion_length": 76.20000076293945,
+      "epoch": 0.057982631930527724,
+      "grad_norm": 11.048291206359863,
+      "kl": 0.0552978515625,
+      "learning_rate": 9.420096205237841e-07,
+      "loss": 0.0022,
+      "reward": 0.39921949803829193,
+      "reward_std": 0.09128798358142376,
+      "rewards/iou_timestamp_reward": 0.39921949803829193,
+      "step": 217
+    },
+    {
+      "completion_length": 83.10000228881836,
+      "epoch": 0.058249832999332,
+      "grad_norm": 14.645618438720703,
+      "kl": 0.069091796875,
+      "learning_rate": 9.417423837520042e-07,
+      "loss": 0.0028,
+      "reward": 0.41773444414138794,
+      "reward_std": 0.06791641190648079,
+      "rewards/iou_timestamp_reward": 0.41773444414138794,
+      "step": 218
+    },
+    {
+      "completion_length": 80.4625015258789,
+      "epoch": 0.05851703406813627,
+      "grad_norm": 13.085001945495605,
+      "kl": 0.0582275390625,
+      "learning_rate": 9.414751469802245e-07,
+      "loss": 0.0023,
+      "reward": 0.44448421895504,
+      "reward_std": 0.07502865791320801,
+      "rewards/iou_timestamp_reward": 0.44448421895504,
+      "step": 219
+    },
+    {
+      "completion_length": 92.125,
+      "epoch": 0.058784235136940546,
+      "grad_norm": 16.1396484375,
+      "kl": 0.0535888671875,
+      "learning_rate": 9.412079102084446e-07,
+      "loss": 0.0021,
+      "reward": 0.48494312167167664,
+      "reward_std": 0.10971744731068611,
+      "rewards/iou_timestamp_reward": 0.48494312167167664,
+      "step": 220
+    },
+    {
+      "completion_length": 87.7750015258789,
+      "epoch": 0.05905143620574482,
+      "grad_norm": 27.641878128051758,
+      "kl": 0.06103515625,
+      "learning_rate": 9.409406734366649e-07,
+      "loss": 0.0024,
+      "reward": 0.566541999578476,
+      "reward_std": 0.07634824328124523,
+      "rewards/iou_timestamp_reward": 0.566541999578476,
+      "step": 221
+    },
+    {
+      "completion_length": 86.26250076293945,
+      "epoch": 0.0593186372745491,
+      "grad_norm": 22.46279525756836,
+      "kl": 0.0477294921875,
+      "learning_rate": 9.406734366648851e-07,
+      "loss": 0.0019,
+      "reward": 0.4909381568431854,
+      "reward_std": 0.060065316036343575,
+      "rewards/iou_timestamp_reward": 0.4909381568431854,
+      "step": 222
+    },
+    {
+      "completion_length": 94.0875015258789,
+      "epoch": 0.059585838343353376,
+      "grad_norm": 16.373811721801758,
+      "kl": 0.0496826171875,
+      "learning_rate": 9.404061998931053e-07,
+      "loss": 0.002,
+      "reward": 0.667538046836853,
+      "reward_std": 0.0789172612130642,
+      "rewards/iou_timestamp_reward": 0.667538046836853,
+      "step": 223
+    },
+    {
+      "completion_length": 87.62500381469727,
+      "epoch": 0.05985303941215765,
+      "grad_norm": 11.457925796508789,
+      "kl": 0.054931640625,
+      "learning_rate": 9.401389631213255e-07,
+      "loss": 0.0022,
+      "reward": 0.523018479347229,
+      "reward_std": 0.11646933481097221,
+      "rewards/iou_timestamp_reward": 0.523018479347229,
+      "step": 224
+    },
+    {
+      "completion_length": 88.86249923706055,
+      "epoch": 0.06012024048096192,
+      "grad_norm": 13.132309913635254,
+      "kl": 0.0538330078125,
+      "learning_rate": 9.398717263495457e-07,
+      "loss": 0.0022,
+      "reward": 0.39982272684574127,
+      "reward_std": 0.09530311450362206,
+      "rewards/iou_timestamp_reward": 0.39982272684574127,
+      "step": 225
+    },
+    {
+      "completion_length": 82.91250228881836,
+      "epoch": 0.0603874415497662,
+      "grad_norm": 13.165536880493164,
+      "kl": 0.0540771484375,
+      "learning_rate": 9.396044895777659e-07,
+      "loss": 0.0022,
+      "reward": 0.4518078416585922,
+      "reward_std": 0.055653566494584084,
+      "rewards/iou_timestamp_reward": 0.4518078416585922,
+      "step": 226
+    },
+    {
+      "completion_length": 87.9124984741211,
+      "epoch": 0.060654642618570474,
+      "grad_norm": 71.02800750732422,
+      "kl": 0.0499267578125,
+      "learning_rate": 9.393372528059861e-07,
+      "loss": 0.002,
+      "reward": 0.4856627434492111,
+      "reward_std": 0.11880143359303474,
+      "rewards/iou_timestamp_reward": 0.4856627434492111,
+      "step": 227
+    },
+    {
+      "completion_length": 88.2125015258789,
+      "epoch": 0.06092184368737475,
+      "grad_norm": 19.84975814819336,
+      "kl": 0.0556640625,
+      "learning_rate": 9.390700160342062e-07,
+      "loss": 0.0022,
+      "reward": 0.5220415741205215,
+      "reward_std": 0.09827179461717606,
+      "rewards/iou_timestamp_reward": 0.5220415741205215,
+      "step": 228
+    },
+    {
+      "completion_length": 85.9375,
+      "epoch": 0.06118904475617903,
+      "grad_norm": 8.188985824584961,
+      "kl": 0.04541015625,
+      "learning_rate": 9.388027792624264e-07,
+      "loss": 0.0018,
+      "reward": 0.4307563751935959,
+      "reward_std": 0.10000541806221008,
+      "rewards/iou_timestamp_reward": 0.4307563751935959,
+      "step": 229
+    },
+    {
+      "completion_length": 83.05000305175781,
+      "epoch": 0.0614562458249833,
+      "grad_norm": 20.375694274902344,
+      "kl": 0.0589599609375,
+      "learning_rate": 9.385355424906467e-07,
+      "loss": 0.0024,
+      "reward": 0.38110922276973724,
+      "reward_std": 0.08979238569736481,
+      "rewards/iou_timestamp_reward": 0.38110922276973724,
+      "step": 230
+    },
+    {
+      "completion_length": 94.28750228881836,
+      "epoch": 0.06172344689378757,
+      "grad_norm": 18.514402389526367,
+      "kl": 0.05224609375,
+      "learning_rate": 9.382683057188668e-07,
+      "loss": 0.0021,
+      "reward": 0.4332323968410492,
+      "reward_std": 0.11731500551104546,
+      "rewards/iou_timestamp_reward": 0.4332323968410492,
+      "step": 231
+    },
+    {
+      "completion_length": 83.4375,
+      "epoch": 0.06199064796259185,
+      "grad_norm": 12.424543380737305,
+      "kl": 0.037841796875,
+      "learning_rate": 9.38001068947087e-07,
+      "loss": 0.0015,
+      "reward": 0.3656405210494995,
+      "reward_std": 0.0752149373292923,
+      "rewards/iou_timestamp_reward": 0.3656405210494995,
+      "step": 232
+    },
+    {
+      "completion_length": 99.26250076293945,
+      "epoch": 0.062257849031396126,
+      "grad_norm": 9.007403373718262,
+      "kl": 0.0421142578125,
+      "learning_rate": 9.377338321753073e-07,
+      "loss": 0.0017,
+      "reward": 0.5581249594688416,
+      "reward_std": 0.10320387035608292,
+      "rewards/iou_timestamp_reward": 0.5581249594688416,
+      "step": 233
+    },
+    {
+      "completion_length": 92.78750228881836,
+      "epoch": 0.0625250501002004,
+      "grad_norm": 19.03189468383789,
+      "kl": 0.062255859375,
+      "learning_rate": 9.374665954035274e-07,
+      "loss": 0.0025,
+      "reward": 0.3949814885854721,
+      "reward_std": 0.11542447470128536,
+      "rewards/iou_timestamp_reward": 0.3949814885854721,
+      "step": 234
+    },
+    {
+      "completion_length": 95.35000228881836,
+      "epoch": 0.06279225116900468,
+      "grad_norm": 39.19453430175781,
+      "kl": 0.0509033203125,
+      "learning_rate": 9.371993586317477e-07,
+      "loss": 0.002,
+      "reward": 0.423181414604187,
+      "reward_std": 0.1289653405547142,
+      "rewards/iou_timestamp_reward": 0.423181414604187,
+      "step": 235
+    },
+    {
+      "completion_length": 98.38750076293945,
+      "epoch": 0.06305945223780896,
+      "grad_norm": 17.67893409729004,
+      "kl": 0.0491943359375,
+      "learning_rate": 9.369321218599678e-07,
+      "loss": 0.002,
+      "reward": 0.5305302441120148,
+      "reward_std": 0.07527987286448479,
+      "rewards/iou_timestamp_reward": 0.5305302441120148,
+      "step": 236
+    },
+    {
+      "completion_length": 89.3125,
+      "epoch": 0.06332665330661323,
+      "grad_norm": 16.596872329711914,
+      "kl": 0.0501708984375,
+      "learning_rate": 9.36664885088188e-07,
+      "loss": 0.002,
+      "reward": 0.47669273614883423,
+      "reward_std": 0.0898175835609436,
+      "rewards/iou_timestamp_reward": 0.47669273614883423,
+      "step": 237
+    },
+    {
+      "completion_length": 80.31250381469727,
+      "epoch": 0.06359385437541751,
+      "grad_norm": 13.407699584960938,
+      "kl": 0.0501708984375,
+      "learning_rate": 9.363976483164083e-07,
+      "loss": 0.002,
+      "reward": 0.48390600085258484,
+      "reward_std": 0.07993911020457745,
+      "rewards/iou_timestamp_reward": 0.48390600085258484,
+      "step": 238
+    },
+    {
+      "completion_length": 89.95000076293945,
+      "epoch": 0.06386105544422177,
+      "grad_norm": 11.182861328125,
+      "kl": 0.0474853515625,
+      "learning_rate": 9.361304115446284e-07,
+      "loss": 0.0019,
+      "reward": 0.5368602573871613,
+      "reward_std": 0.08251865021884441,
+      "rewards/iou_timestamp_reward": 0.5368602573871613,
+      "step": 239
+    },
+    {
+      "completion_length": 90.91250228881836,
+      "epoch": 0.06412825651302605,
+      "grad_norm": 17.626962661743164,
+      "kl": 0.0478515625,
+      "learning_rate": 9.358631747728487e-07,
+      "loss": 0.0019,
+      "reward": 0.589008092880249,
+      "reward_std": 0.14458639919757843,
+      "rewards/iou_timestamp_reward": 0.589008092880249,
+      "step": 240
+    },
+    {
+      "completion_length": 94.0250015258789,
+      "epoch": 0.06439545758183032,
+      "grad_norm": 19.81915283203125,
+      "kl": 0.0458984375,
+      "learning_rate": 9.355959380010689e-07,
+      "loss": 0.0018,
+      "reward": 0.38526751101017,
+      "reward_std": 0.1254762001335621,
+      "rewards/iou_timestamp_reward": 0.38526751101017,
+      "step": 241
+    },
+    {
+      "completion_length": 90.56250381469727,
+      "epoch": 0.0646626586506346,
+      "grad_norm": 13.259626388549805,
+      "kl": 0.05224609375,
+      "learning_rate": 9.353287012292891e-07,
+      "loss": 0.0021,
+      "reward": 0.3110887184739113,
+      "reward_std": 0.1004025787115097,
+      "rewards/iou_timestamp_reward": 0.3110887184739113,
+      "step": 242
+    },
+    {
+      "completion_length": 90.7125015258789,
+      "epoch": 0.06492985971943888,
+      "grad_norm": 43.69268035888672,
+      "kl": 0.0465087890625,
+      "learning_rate": 9.350614644575093e-07,
+      "loss": 0.0019,
+      "reward": 0.5469205528497696,
+      "reward_std": 0.09666259959340096,
+      "rewards/iou_timestamp_reward": 0.5469205528497696,
+      "step": 243
+    },
+    {
+      "completion_length": 91.23749923706055,
+      "epoch": 0.06519706078824315,
+      "grad_norm": 36.97768783569336,
+      "kl": 0.0537109375,
+      "learning_rate": 9.347942276857294e-07,
+      "loss": 0.0022,
+      "reward": 0.45238882303237915,
+      "reward_std": 0.13411134481430054,
+      "rewards/iou_timestamp_reward": 0.45238882303237915,
+      "step": 244
+    },
+    {
+      "completion_length": 90.42499923706055,
+      "epoch": 0.06546426185704743,
+      "grad_norm": 13.66786003112793,
+      "kl": 0.05322265625,
+      "learning_rate": 9.345269909139497e-07,
+      "loss": 0.0021,
+      "reward": 0.3465211093425751,
+      "reward_std": 0.06062236800789833,
+      "rewards/iou_timestamp_reward": 0.3465211093425751,
+      "step": 245
+    },
+    {
+      "completion_length": 88.97500228881836,
+      "epoch": 0.0657314629258517,
+      "grad_norm": 37.682861328125,
+      "kl": 0.05517578125,
+      "learning_rate": 9.342597541421699e-07,
+      "loss": 0.0022,
+      "reward": 0.6093435287475586,
+      "reward_std": 0.1017681360244751,
+      "rewards/iou_timestamp_reward": 0.6093435287475586,
+      "step": 246
+    },
+    {
+      "completion_length": 89.5875015258789,
+      "epoch": 0.06599866399465598,
+      "grad_norm": 16.518226623535156,
+      "kl": 0.0675048828125,
+      "learning_rate": 9.339925173703901e-07,
+      "loss": 0.0027,
+      "reward": 0.372649610042572,
+      "reward_std": 0.09408055618405342,
+      "rewards/iou_timestamp_reward": 0.372649610042572,
+      "step": 247
+    },
+    {
+      "completion_length": 96.05000305175781,
+      "epoch": 0.06626586506346026,
+      "grad_norm": 15.585586547851562,
+      "kl": 0.0504150390625,
+      "learning_rate": 9.337252805986103e-07,
+      "loss": 0.002,
+      "reward": 0.4263141304254532,
+      "reward_std": 0.09799651429057121,
+      "rewards/iou_timestamp_reward": 0.4263141304254532,
+      "step": 248
+    },
+    {
+      "completion_length": 95.10000228881836,
+      "epoch": 0.06653306613226453,
+      "grad_norm": 19.904935836791992,
+      "kl": 0.0557861328125,
+      "learning_rate": 9.334580438268306e-07,
+      "loss": 0.0022,
+      "reward": 0.49050794541835785,
+      "reward_std": 0.11211569979786873,
+      "rewards/iou_timestamp_reward": 0.49050794541835785,
+      "step": 249
+    },
+    {
+      "completion_length": 92.03750228881836,
+      "epoch": 0.06680026720106881,
+      "grad_norm": 357.7543640136719,
+      "kl": 0.0576171875,
+      "learning_rate": 9.331908070550507e-07,
+      "loss": 0.0023,
+      "reward": 0.411429226398468,
+      "reward_std": 0.15467696636915207,
+      "rewards/iou_timestamp_reward": 0.411429226398468,
+      "step": 250
+    },
+    {
+      "completion_length": 92.78750228881836,
+      "epoch": 0.06706746826987307,
+      "grad_norm": 22.54250717163086,
+      "kl": 0.0506591796875,
+      "learning_rate": 9.32923570283271e-07,
+      "loss": 0.002,
+      "reward": 0.5628893673419952,
+      "reward_std": 0.10398486256599426,
+      "rewards/iou_timestamp_reward": 0.5628893673419952,
+      "step": 251
+    },
+    {
+      "completion_length": 87.8375015258789,
+      "epoch": 0.06733466933867735,
+      "grad_norm": 37.64324951171875,
+      "kl": 0.04833984375,
+      "learning_rate": 9.326563335114911e-07,
+      "loss": 0.0019,
+      "reward": 0.47143837809562683,
+      "reward_std": 0.1283393558114767,
+      "rewards/iou_timestamp_reward": 0.47143837809562683,
+      "step": 252
+    },
+    {
+      "completion_length": 90.7750015258789,
+      "epoch": 0.06760187040748163,
+      "grad_norm": 11.761100769042969,
+      "kl": 0.0535888671875,
+      "learning_rate": 9.323890967397113e-07,
+      "loss": 0.0021,
+      "reward": 0.4104880392551422,
+      "reward_std": 0.12630710378289223,
+      "rewards/iou_timestamp_reward": 0.4104880392551422,
+      "step": 253
+    },
+    {
+      "completion_length": 89.51250076293945,
+      "epoch": 0.0678690714762859,
+      "grad_norm": 17.23011016845703,
+      "kl": 0.051025390625,
+      "learning_rate": 9.321218599679316e-07,
+      "loss": 0.002,
+      "reward": 0.43392352759838104,
+      "reward_std": 0.11164272576570511,
+      "rewards/iou_timestamp_reward": 0.43392352759838104,
+      "step": 254
+    },
+    {
+      "completion_length": 94.2125015258789,
+      "epoch": 0.06813627254509018,
+      "grad_norm": 20.68499183654785,
+      "kl": 0.056640625,
+      "learning_rate": 9.318546231961517e-07,
+      "loss": 0.0023,
+      "reward": 0.4146452210843563,
+      "reward_std": 0.080996323376894,
+      "rewards/iou_timestamp_reward": 0.4146452210843563,
+      "step": 255
+    },
+    {
+      "completion_length": 83.43750381469727,
+      "epoch": 0.06840347361389446,
+      "grad_norm": 21.620370864868164,
+      "kl": 0.040283203125,
+      "learning_rate": 9.31587386424372e-07,
+      "loss": 0.0016,
+      "reward": 0.5084521472454071,
+      "reward_std": 0.11804941296577454,
+      "rewards/iou_timestamp_reward": 0.5084521472454071,
+      "step": 256
+    },
+    {
+      "completion_length": 86.13750076293945,
+      "epoch": 0.06867067468269873,
+      "grad_norm": 35.96171569824219,
+      "kl": 0.0594482421875,
+      "learning_rate": 9.313201496525922e-07,
+      "loss": 0.0024,
+      "reward": 0.41805142164230347,
+      "reward_std": 0.10501766204833984,
+      "rewards/iou_timestamp_reward": 0.41805142164230347,
+      "step": 257
+    },
+    {
+      "completion_length": 84.26250457763672,
+      "epoch": 0.06893787575150301,
+      "grad_norm": 13.32675552368164,
+      "kl": 0.05615234375,
+      "learning_rate": 9.310529128808123e-07,
+      "loss": 0.0022,
+      "reward": 0.3628458231687546,
+      "reward_std": 0.12849557399749756,
+      "rewards/iou_timestamp_reward": 0.3628458231687546,
+      "step": 258
+    },
+    {
+      "completion_length": 85.36249923706055,
+      "epoch": 0.06920507682030728,
+      "grad_norm": 10.719467163085938,
+      "kl": 0.057861328125,
+      "learning_rate": 9.307856761090326e-07,
+      "loss": 0.0023,
+      "reward": 0.375208780169487,
+      "reward_std": 0.07890206202864647,
+      "rewards/iou_timestamp_reward": 0.375208780169487,
+      "step": 259
+    },
+    {
+      "completion_length": 88.38750076293945,
+      "epoch": 0.06947227788911156,
+      "grad_norm": 16.24102783203125,
+      "kl": 0.0723876953125,
+      "learning_rate": 9.305184393372527e-07,
+      "loss": 0.0029,
+      "reward": 0.46249307692050934,
+      "reward_std": 0.0996074341237545,
+      "rewards/iou_timestamp_reward": 0.46249307692050934,
+      "step": 260
+    },
+    {
+      "completion_length": 84.4625015258789,
+      "epoch": 0.06973947895791584,
+      "grad_norm": 10.950820922851562,
+      "kl": 0.05029296875,
+      "learning_rate": 9.30251202565473e-07,
+      "loss": 0.002,
+      "reward": 0.43129654228687286,
+      "reward_std": 0.09118984453380108,
+      "rewards/iou_timestamp_reward": 0.43129654228687286,
+      "step": 261
+    },
+    {
+      "completion_length": 78.35000228881836,
+      "epoch": 0.0700066800267201,
+      "grad_norm": 23.676605224609375,
+      "kl": 0.0572509765625,
+      "learning_rate": 9.299839657936932e-07,
+      "loss": 0.0023,
+      "reward": 0.4307858943939209,
+      "reward_std": 0.0822877585887909,
+      "rewards/iou_timestamp_reward": 0.4307858943939209,
+      "step": 262
+    },
+    {
+      "completion_length": 84.36250305175781,
+      "epoch": 0.07027388109552438,
+      "grad_norm": 30.50290870666504,
+      "kl": 0.050537109375,
+      "learning_rate": 9.297167290219134e-07,
+      "loss": 0.002,
+      "reward": 0.4225500524044037,
+      "reward_std": 0.08897064253687859,
+      "rewards/iou_timestamp_reward": 0.4225500524044037,
+      "step": 263
+    },
+    {
+      "completion_length": 86.12500381469727,
+      "epoch": 0.07054108216432865,
+      "grad_norm": 9.197006225585938,
+      "kl": 0.046142578125,
+      "learning_rate": 9.294494922501336e-07,
+      "loss": 0.0018,
+      "reward": 0.523790642619133,
+      "reward_std": 0.11769257113337517,
+      "rewards/iou_timestamp_reward": 0.523790642619133,
+      "step": 264
+    },
+    {
+      "completion_length": 85.04999923706055,
+      "epoch": 0.07080828323313293,
+      "grad_norm": 30.433368682861328,
+      "kl": 0.0567626953125,
+      "learning_rate": 9.291822554783539e-07,
+      "loss": 0.0023,
+      "reward": 0.3194416016340256,
+      "reward_std": 0.07132554426789284,
+      "rewards/iou_timestamp_reward": 0.3194416016340256,
+      "step": 265
+    },
+    {
+      "completion_length": 84.7249984741211,
+      "epoch": 0.0710754843019372,
+      "grad_norm": 24.936206817626953,
+      "kl": 0.0491943359375,
+      "learning_rate": 9.28915018706574e-07,
+      "loss": 0.002,
+      "reward": 0.37983009219169617,
+      "reward_std": 0.08205056935548782,
+      "rewards/iou_timestamp_reward": 0.37983009219169617,
+      "step": 266
+    },
+    {
+      "completion_length": 82.23750305175781,
+      "epoch": 0.07134268537074148,
+      "grad_norm": 36.05076217651367,
+      "kl": 0.046630859375,
+      "learning_rate": 9.286477819347942e-07,
+      "loss": 0.0019,
+      "reward": 0.5814423114061356,
+      "reward_std": 0.08882495388388634,
+      "rewards/iou_timestamp_reward": 0.5814423114061356,
+      "step": 267
+    },
+    {
+      "completion_length": 85.0625,
+      "epoch": 0.07160988643954576,
+      "grad_norm": 15.191946983337402,
+      "kl": 0.0489501953125,
+      "learning_rate": 9.283805451630144e-07,
+      "loss": 0.002,
+      "reward": 0.4915848672389984,
+      "reward_std": 0.08231451734900475,
+      "rewards/iou_timestamp_reward": 0.4915848672389984,
+      "step": 268
+    },
+    {
+      "completion_length": 87.07500076293945,
+      "epoch": 0.07187708750835003,
+      "grad_norm": 12.557903289794922,
+      "kl": 0.057861328125,
+      "learning_rate": 9.281133083912346e-07,
+      "loss": 0.0023,
+      "reward": 0.5351927578449249,
+      "reward_std": 0.06344357319176197,
+      "rewards/iou_timestamp_reward": 0.5351927578449249,
+      "step": 269
+    },
+    {
+      "completion_length": 84.0875015258789,
+      "epoch": 0.07214428857715431,
+      "grad_norm": 14.385464668273926,
+      "kl": 0.0570068359375,
+      "learning_rate": 9.278460716194549e-07,
+      "loss": 0.0023,
+      "reward": 0.4179234653711319,
+      "reward_std": 0.10309569910168648,
+      "rewards/iou_timestamp_reward": 0.4179234653711319,
+      "step": 270
+    },
+    {
+      "completion_length": 89.0374984741211,
+      "epoch": 0.07241148964595859,
+      "grad_norm": 9.917317390441895,
+      "kl": 0.0496826171875,
+      "learning_rate": 9.27578834847675e-07,
+      "loss": 0.002,
+      "reward": 0.3758576214313507,
+      "reward_std": 0.06827409937977791,
+      "rewards/iou_timestamp_reward": 0.3758576214313507,
+      "step": 271
+    },
+    {
+      "completion_length": 94.4625015258789,
+      "epoch": 0.07267869071476286,
+      "grad_norm": 24.282136917114258,
+      "kl": 0.048828125,
+      "learning_rate": 9.273115980758952e-07,
+      "loss": 0.002,
+      "reward": 0.472480833530426,
+      "reward_std": 0.055225247517228127,
+      "rewards/iou_timestamp_reward": 0.472480833530426,
+      "step": 272
+    },
+    {
+      "completion_length": 86.7125015258789,
+      "epoch": 0.07294589178356714,
+      "grad_norm": 11.784228324890137,
+      "kl": 0.05126953125,
+      "learning_rate": 9.270443613041155e-07,
+      "loss": 0.0021,
+      "reward": 0.505023866891861,
+      "reward_std": 0.09757990017533302,
+      "rewards/iou_timestamp_reward": 0.505023866891861,
+      "step": 273
+    },
+    {
+      "completion_length": 89.78750228881836,
+      "epoch": 0.0732130928523714,
+      "grad_norm": 45.109092712402344,
+      "kl": 0.050048828125,
+      "learning_rate": 9.267771245323356e-07,
+      "loss": 0.002,
+      "reward": 0.523750826716423,
+      "reward_std": 0.17876286804676056,
+      "rewards/iou_timestamp_reward": 0.523750826716423,
+      "step": 274
+    },
+    {
+      "completion_length": 93.31250381469727,
+      "epoch": 0.07348029392117568,
+      "grad_norm": 14.207196235656738,
+      "kl": 0.048828125,
+      "learning_rate": 9.265098877605559e-07,
+      "loss": 0.002,
+      "reward": 0.4920515716075897,
+      "reward_std": 0.09994343668222427,
+      "rewards/iou_timestamp_reward": 0.4920515716075897,
+      "step": 275
+    },
+    {
+      "completion_length": 92.17500305175781,
+      "epoch": 0.07374749498997996,
+      "grad_norm": 48.04758834838867,
+      "kl": 0.0509033203125,
+      "learning_rate": 9.26242650988776e-07,
+      "loss": 0.002,
+      "reward": 0.4091592878103256,
+      "reward_std": 0.07305929809808731,
+      "rewards/iou_timestamp_reward": 0.4091592878103256,
+      "step": 276
+    },
+    {
+      "completion_length": 88.2750015258789,
+      "epoch": 0.07401469605878423,
+      "grad_norm": 11.50412654876709,
+      "kl": 0.059326171875,
+      "learning_rate": 9.259754142169963e-07,
+      "loss": 0.0024,
+      "reward": 0.43483787775039673,
+      "reward_std": 0.13717280700802803,
+      "rewards/iou_timestamp_reward": 0.43483787775039673,
+      "step": 277
+    },
+    {
+      "completion_length": 92.61250305175781,
+      "epoch": 0.07428189712758851,
+      "grad_norm": 37.55918884277344,
+      "kl": 0.051513671875,
+      "learning_rate": 9.257081774452165e-07,
+      "loss": 0.0021,
+      "reward": 0.3264332413673401,
+      "reward_std": 0.07515481114387512,
+      "rewards/iou_timestamp_reward": 0.3264332413673401,
+      "step": 278
+    },
+    {
+      "completion_length": 87.92500305175781,
+      "epoch": 0.07454909819639279,
+      "grad_norm": 15.809645652770996,
+      "kl": 0.0555419921875,
+      "learning_rate": 9.254409406734365e-07,
+      "loss": 0.0022,
+      "reward": 0.4322213977575302,
+      "reward_std": 0.0831381669268012,
+      "rewards/iou_timestamp_reward": 0.4322213977575302,
+      "step": 279
+    },
+    {
+      "completion_length": 88.82500076293945,
+      "epoch": 0.07481629926519706,
+      "grad_norm": 29.246431350708008,
+      "kl": 0.0552978515625,
+      "learning_rate": 9.251737039016568e-07,
+      "loss": 0.0022,
+      "reward": 0.6085447072982788,
+      "reward_std": 0.13383636623620987,
+      "rewards/iou_timestamp_reward": 0.6085447072982788,
+      "step": 280
+    },
+    {
+      "completion_length": 90.7750015258789,
+      "epoch": 0.07508350033400134,
+      "grad_norm": 56.85616683959961,
+      "kl": 0.05810546875,
+      "learning_rate": 9.24906467129877e-07,
+      "loss": 0.0023,
+      "reward": 0.5065873116254807,
+      "reward_std": 0.09172694198787212,
+      "rewards/iou_timestamp_reward": 0.5065873116254807,
+      "step": 281
+    },
+    {
+      "completion_length": 94.125,
+      "epoch": 0.07535070140280561,
+      "grad_norm": 153.61770629882812,
+      "kl": 0.0584716796875,
+      "learning_rate": 9.246392303580972e-07,
+      "loss": 0.0023,
+      "reward": 0.619166225194931,
+      "reward_std": 0.08932790905237198,
+      "rewards/iou_timestamp_reward": 0.619166225194931,
+      "step": 282
+    },
+    {
+      "completion_length": 89.7750015258789,
+      "epoch": 0.07561790247160989,
+      "grad_norm": 92.22215270996094,
+      "kl": 0.0556640625,
+      "learning_rate": 9.243719935863174e-07,
+      "loss": 0.0022,
+      "reward": 0.5364521145820618,
+      "reward_std": 0.0949009507894516,
+      "rewards/iou_timestamp_reward": 0.5364521145820618,
+      "step": 283
+    },
+    {
+      "completion_length": 90.85000228881836,
+      "epoch": 0.07588510354041417,
+      "grad_norm": 11.400144577026367,
+      "kl": 0.0491943359375,
+      "learning_rate": 9.241047568145376e-07,
+      "loss": 0.002,
+      "reward": 0.3629377633333206,
+      "reward_std": 0.051846086978912354,
+      "rewards/iou_timestamp_reward": 0.3629377633333206,
+      "step": 284
+    },
+    {
+      "completion_length": 88.05000305175781,
+      "epoch": 0.07615230460921844,
+      "grad_norm": 10.160490036010742,
+      "kl": 0.0489501953125,
+      "learning_rate": 9.238375200427578e-07,
+      "loss": 0.002,
+      "reward": 0.4853304475545883,
+      "reward_std": 0.0822471622377634,
+      "rewards/iou_timestamp_reward": 0.4853304475545883,
+      "step": 285
+    },
+    {
+      "completion_length": 83.75,
+      "epoch": 0.0764195056780227,
+      "grad_norm": 29.11613655090332,
+      "kl": 0.0517578125,
+      "learning_rate": 9.23570283270978e-07,
+      "loss": 0.0021,
+      "reward": 0.5097087770700455,
+      "reward_std": 0.0779729075729847,
+      "rewards/iou_timestamp_reward": 0.5097087770700455,
+      "step": 286
+    },
+    {
+      "completion_length": 93.97500228881836,
+      "epoch": 0.07668670674682698,
+      "grad_norm": 13.363306045532227,
+      "kl": 0.0643310546875,
+      "learning_rate": 9.233030464991982e-07,
+      "loss": 0.0026,
+      "reward": 0.3735619783401489,
+      "reward_std": 0.12772585824131966,
+      "rewards/iou_timestamp_reward": 0.3735619783401489,
+      "step": 287
+    },
+    {
+      "completion_length": 88.7125015258789,
+      "epoch": 0.07695390781563126,
+      "grad_norm": 253.6396026611328,
+      "kl": 0.0447998046875,
+      "learning_rate": 9.230358097274184e-07,
+      "loss": 0.0018,
+      "reward": 0.5121918767690659,
+      "reward_std": 0.11110168322920799,
+      "rewards/iou_timestamp_reward": 0.5121918767690659,
+      "step": 288
+    },
+    {
+      "completion_length": 87.95000076293945,
+      "epoch": 0.07722110888443554,
+      "grad_norm": 9.944025039672852,
+      "kl": 0.0465087890625,
+      "learning_rate": 9.227685729556387e-07,
+      "loss": 0.0019,
+      "reward": 0.5837224572896957,
+      "reward_std": 0.0957045927643776,
+      "rewards/iou_timestamp_reward": 0.5837224572896957,
+      "step": 289
+    },
+    {
+      "completion_length": 88.3375015258789,
+      "epoch": 0.07748830995323981,
+      "grad_norm": 19.37244415283203,
+      "kl": 0.0513916015625,
+      "learning_rate": 9.225013361838588e-07,
+      "loss": 0.0021,
+      "reward": 0.4252195358276367,
+      "reward_std": 0.13499490171670914,
+      "rewards/iou_timestamp_reward": 0.4252195358276367,
+      "step": 290
+    },
+    {
+      "completion_length": 90.23749923706055,
+      "epoch": 0.07775551102204409,
+      "grad_norm": 13.233763694763184,
+      "kl": 0.050048828125,
+      "learning_rate": 9.222340994120791e-07,
+      "loss": 0.002,
+      "reward": 0.4406256824731827,
+      "reward_std": 0.10843142122030258,
+      "rewards/iou_timestamp_reward": 0.4406256824731827,
+      "step": 291
+    },
+    {
+      "completion_length": 94.42499923706055,
+      "epoch": 0.07802271209084836,
+      "grad_norm": 13.899524688720703,
+      "kl": 0.05615234375,
+      "learning_rate": 9.219668626402992e-07,
+      "loss": 0.0022,
+      "reward": 0.4787358343601227,
+      "reward_std": 0.13102044723927975,
+      "rewards/iou_timestamp_reward": 0.4787358343601227,
+      "step": 292
+    },
+    {
+      "completion_length": 86.04999923706055,
+      "epoch": 0.07828991315965264,
+      "grad_norm": 12.498344421386719,
+      "kl": 0.048583984375,
+      "learning_rate": 9.216996258685194e-07,
+      "loss": 0.0019,
+      "reward": 0.4275331050157547,
+      "reward_std": 0.1066926084458828,
+      "rewards/iou_timestamp_reward": 0.4275331050157547,
+      "step": 293
+    },
+    {
+      "completion_length": 90.42499923706055,
+      "epoch": 0.07855711422845692,
+      "grad_norm": 22.13373565673828,
+      "kl": 0.0589599609375,
+      "learning_rate": 9.214323890967397e-07,
+      "loss": 0.0024,
+      "reward": 0.35555990040302277,
+      "reward_std": 0.09240290522575378,
+      "rewards/iou_timestamp_reward": 0.35555990040302277,
+      "step": 294
+    },
+    {
+      "completion_length": 91.51250457763672,
+      "epoch": 0.0788243152972612,
+      "grad_norm": 44.67702865600586,
+      "kl": 0.052490234375,
+      "learning_rate": 9.211651523249598e-07,
+      "loss": 0.0021,
+      "reward": 0.41058990359306335,
+      "reward_std": 0.09301923587918282,
+      "rewards/iou_timestamp_reward": 0.41058990359306335,
+      "step": 295
+    },
+    {
+      "completion_length": 88.375,
+      "epoch": 0.07909151636606547,
+      "grad_norm": 16.11538314819336,
+      "kl": 0.065185546875,
+      "learning_rate": 9.208979155531801e-07,
+      "loss": 0.0026,
+      "reward": 0.3699526935815811,
+      "reward_std": 0.08414144068956375,
+      "rewards/iou_timestamp_reward": 0.3699526935815811,
+      "step": 296
+    },
+    {
+      "completion_length": 87.55000305175781,
+      "epoch": 0.07935871743486973,
+      "grad_norm": 74.45164489746094,
+      "kl": 0.04736328125,
+      "learning_rate": 9.206306787814003e-07,
+      "loss": 0.0019,
+      "reward": 0.5554560720920563,
+      "reward_std": 0.11732692644000053,
+      "rewards/iou_timestamp_reward": 0.5554560720920563,
+      "step": 297
+    },
+    {
+      "completion_length": 84.18750381469727,
+      "epoch": 0.07962591850367401,
+      "grad_norm": 107.87571716308594,
+      "kl": 0.04345703125,
+      "learning_rate": 9.203634420096205e-07,
+      "loss": 0.0017,
+      "reward": 0.3576356768608093,
+      "reward_std": 0.05249842070043087,
+      "rewards/iou_timestamp_reward": 0.3576356768608093,
+      "step": 298
+    },
+    {
+      "completion_length": 80.6875,
+      "epoch": 0.07989311957247829,
+      "grad_norm": 16.578479766845703,
+      "kl": 0.0604248046875,
+      "learning_rate": 9.200962052378407e-07,
+      "loss": 0.0024,
+      "reward": 0.5808127671480179,
+      "reward_std": 0.09874389320611954,
+      "rewards/iou_timestamp_reward": 0.5808127671480179,
+      "step": 299
+    },
+    {
+      "completion_length": 82.1500015258789,
+      "epoch": 0.08016032064128256,
+      "grad_norm": 25.647584915161133,
+      "kl": 0.071044921875,
+      "learning_rate": 9.198289684660608e-07,
+      "loss": 0.0028,
+      "reward": 0.4414636939764023,
+      "reward_std": 0.0884641632437706,
+      "rewards/iou_timestamp_reward": 0.4414636939764023,
+      "step": 300
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3742,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}