Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -4687,3 +4687,59 @@
 {"timestamp": 1787544794.5421376, "elapsed_hours": 5.047403786248631, "session_id": 2, "event": "train_step", "step": 41500, "epoch": 1, "loss": 1.9144518899917602, "loss_raw": 2.7311737537384033, "perplexity": 6.783219827214752, "muon_lr": 0.008509075907554967, "adamw_lr": 0.00025527227722664894, "tokens_per_sec": 149770.46066226828, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2719744000, "tokens_processed_billions": 2.719744}
 {"timestamp": 1787544794.5453146, "elapsed_hours": 5.04740466872851, "session_id": 2, "event": "lambda_stats", "step": 41500, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1896701388888888, "lambda_std": 0.2043699421791281, "lambda_all": [2.015625, 1.2421875, 1.203125, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
 {"timestamp": 1787544808.0766318, "elapsed_hours": 5.051163368291325, "session_id": 2, "event": "validation", "step": 41500, "epoch": 1, "val_loss": 1.9705719661712646, "val_perplexity": 7.1747790458732466, "is_best": true, "best_loss": 1.9705719661712646, "best_perplexity": 7.1747790458732466, "train_loss": 2.7311737537384033, "train_loss_avg": 1.9144518899917602}
+{"timestamp": 1787544812.9937925, "elapsed_hours": 5.052529245879915, "session_id": 2, "event": "train_step", "step": 41510, "epoch": 1, "loss": 1.9587551867961883, "loss_raw": 1.2571519613265991, "perplexity": 7.090495227222441, "muon_lr": 0.008508507666880858, "adamw_lr": 0.00025525523000642566, "tokens_per_sec": 149654.4867991075, "vram_gb": 27.710740089416504, "grad_norm": 0.1396484375, "tokens_processed": 2720399360, "tokens_processed_billions": 2.72039936}
+{"timestamp": 1787544817.228637, "elapsed_hours": 5.053705591691865, "session_id": 2, "event": "train_step", "step": 41520, "epoch": 1, "loss": 1.9668532288074494, "loss_raw": 1.9164824485778809, "perplexity": 7.14814747555671, "muon_lr": 0.0085073709252763, "adamw_lr": 0.00025522112775828895, "tokens_per_sec": 149655.67477602893, "vram_gb": 27.299458026885986, "grad_norm": 0.1640625, "tokens_processed": 2721054720, "tokens_processed_billions": 2.72105472}
+{"timestamp": 1787544821.2607744, "elapsed_hours": 5.054825629856851, "session_id": 2, "event": "train_step", "step": 41530, "epoch": 1, "loss": 2.047690500020981, "loss_raw": 4.044757843017578, "perplexity": 7.749981839103149, "muon_lr": 0.008506802424378634, "adamw_lr": 0.00025520407273135894, "tokens_per_sec": 149658.53026618625, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 2721710080, "tokens_processed_billions": 2.72171008}
+{"timestamp": 1787544825.270563, "elapsed_hours": 5.055939459866948, "session_id": 2, "event": "train_step", "step": 41540, "epoch": 1, "loss": 2.0941927671432494, "loss_raw": 2.773852825164795, "perplexity": 8.11888449394345, "muon_lr": 0.008506233836772563, "adamw_lr": 0.00025518701510317683, "tokens_per_sec": 149661.56896655867, "vram_gb": 27.710740089416504, "grad_norm": 0.15625, "tokens_processed": 2722365440, "tokens_processed_billions": 2.72236544}
+{"timestamp": 1787544829.343783, "elapsed_hours": 5.057070909937223, "session_id": 2, "event": "train_step", "step": 41550, "epoch": 1, "loss": 2.0358977460861207, "loss_raw": 2.600755214691162, "perplexity": 7.6591249894560685, "muon_lr": 0.00850566516247449, "adamw_lr": 0.0002551699548742346, "tokens_per_sec": 149664.0833472602, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2723020800, "tokens_processed_billions": 2.7230208}
+{"timestamp": 1787544833.4900556, "elapsed_hours": 5.058222652408812, "session_id": 2, "event": "train_step", "step": 41560, "epoch": 1, "loss": 2.0582767760753633, "loss_raw": 0.9224407076835632, "perplexity": 7.8324610900471585, "muon_lr": 0.008504527553867927, "adamw_lr": 0.00025513582661603775, "tokens_per_sec": 149665.9961063237, "vram_gb": 27.299458026885986, "grad_norm": 0.171875, "tokens_processed": 2723676160, "tokens_processed_billions": 2.72367616}
+{"timestamp": 1787544837.6019719, "elapsed_hours": 5.0593648513820435, "session_id": 2, "event": "train_step", "step": 41570, "epoch": 1, "loss": 2.02462819814682, "loss_raw": 2.156463384628296, "perplexity": 7.57329465613082, "muon_lr": 0.008503958619592241, "adamw_lr": 0.0002551187585877672, "tokens_per_sec": 149668.19126567908, "vram_gb": 27.710740089416504, "grad_norm": 0.15234375, "tokens_processed": 2724331520, "tokens_processed_billions": 2.72433152}
+{"timestamp": 1787544841.7002437, "elapsed_hours": 5.060503260625733, "session_id": 2, "event": "train_step", "step": 41580, "epoch": 1, "loss": 1.9558860230445863, "loss_raw": 1.1721066236495972, "perplexity": 7.070180592278698, "muon_lr": 0.008503389598690165, "adamw_lr": 0.0002551016879607049, "tokens_per_sec": 149670.4963125062, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 2724986880, "tokens_processed_billions": 2.72498688}
+{"timestamp": 1787544845.8138394, "elapsed_hours": 5.061645925773514, "session_id": 2, "event": "train_step", "step": 41590, "epoch": 1, "loss": 1.942333209514618, "loss_raw": 3.1210224628448486, "perplexity": 6.975006150607853, "muon_lr": 0.008502820491178105, "adamw_lr": 0.0002550846147353431, "tokens_per_sec": 149672.67529460668, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 2725642240, "tokens_processed_billions": 2.72564224}
+{"timestamp": 1787544850.0539474, "elapsed_hours": 5.0628237334224915, "session_id": 2, "event": "train_step", "step": 41600, "epoch": 1, "loss": 2.0359941387176512, "loss_raw": 2.8475255966186523, "perplexity": 7.659863308252697, "muon_lr": 0.008501682016389678, "adamw_lr": 0.0002550504604916903, "tokens_per_sec": 149673.81291249846, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 2726297600, "tokens_processed_billions": 2.7262976}
+{"timestamp": 1787544850.057511, "elapsed_hours": 5.062824723521868, "session_id": 2, "event": "lambda_stats", "step": 41600, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1896701388888888, "lambda_std": 0.2043699421791281, "lambda_all": [2.015625, 1.2421875, 1.203125, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787544854.0333211, "elapsed_hours": 5.063929114871555, "session_id": 2, "event": "train_step", "step": 41610, "epoch": 1, "loss": 1.9586531114578247, "loss_raw": 1.3388240337371826, "perplexity": 7.089771499460963, "muon_lr": 0.008501112649146143, "adamw_lr": 0.00025503337947438426, "tokens_per_sec": 149677.09286068048, "vram_gb": 27.710740089416504, "grad_norm": 0.1318359375, "tokens_processed": 2726952960, "tokens_processed_billions": 2.72695296}
+{"timestamp": 1787544858.0783043, "elapsed_hours": 5.065052721301715, "session_id": 2, "event": "train_step", "step": 41620, "epoch": 1, "loss": 1.902434356212616, "loss_raw": 1.6734739542007446, "perplexity": 6.702190117643538, "muon_lr": 0.008500543195358286, "adamw_lr": 0.0002550162958607485, "tokens_per_sec": 149679.8322339812, "vram_gb": 27.710740089416504, "grad_norm": 0.14453125, "tokens_processed": 2727608320, "tokens_processed_billions": 2.72760832}
+{"timestamp": 1787544862.1885178, "elapsed_hours": 5.066194447543886, "session_id": 2, "event": "train_step", "step": 41630, "epoch": 1, "loss": 1.8578850972652434, "loss_raw": 1.4785798788070679, "perplexity": 6.410165548930133, "muon_lr": 0.008499973655042525, "adamw_lr": 0.0002549992096512757, "tokens_per_sec": 149682.0349084797, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 2728263680, "tokens_processed_billions": 2.72826368}
+{"timestamp": 1787544866.3322256, "elapsed_hours": 5.067345477342606, "session_id": 2, "event": "train_step", "step": 41640, "epoch": 1, "loss": 1.7591148936748504, "loss_raw": 1.4441801309585571, "perplexity": 5.807295045398797, "muon_lr": 0.008498834314893003, "adamw_lr": 0.00025496502944679, "tokens_per_sec": 149683.9601973062, "vram_gb": 27.299458026885986, "grad_norm": 0.15234375, "tokens_processed": 2728919040, "tokens_processed_billions": 2.72891904}
+{"timestamp": 1787544870.3767638, "elapsed_hours": 5.068468960192469, "session_id": 2, "event": "train_step", "step": 41650, "epoch": 1, "loss": 1.815704151391983, "loss_raw": 1.5407291650772095, "perplexity": 6.145401947603363, "muon_lr": 0.008498264515092092, "adamw_lr": 0.0002549479354527627, "tokens_per_sec": 149686.70082820015, "vram_gb": 27.710740089416504, "grad_norm": 0.16015625, "tokens_processed": 2729574400, "tokens_processed_billions": 2.7295744}
+{"timestamp": 1787544874.3704352, "elapsed_hours": 5.069578313363923, "session_id": 2, "event": "train_step", "step": 41660, "epoch": 1, "loss": 1.8999935817718505, "loss_raw": 2.8027477264404297, "perplexity": 6.685851530821063, "muon_lr": 0.008497694628828994, "adamw_lr": 0.00025493083886486977, "tokens_per_sec": 149689.85693094504, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 2730229760, "tokens_processed_billions": 2.73022976}
+{"timestamp": 1787544878.357063, "elapsed_hours": 5.070685709979799, "session_id": 2, "event": "train_step", "step": 41670, "epoch": 1, "loss": 1.9240861070156097, "loss_raw": 2.5063302516937256, "perplexity": 6.848886655465998, "muon_lr": 0.008497124656120139, "adamw_lr": 0.0002549137396836041, "tokens_per_sec": 149693.06873142684, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 2730885120, "tokens_processed_billions": 2.73088512}
+{"timestamp": 1787544882.4824538, "elapsed_hours": 5.071831651793586, "session_id": 2, "event": "train_step", "step": 41680, "epoch": 1, "loss": 2.0120757389068604, "loss_raw": 1.779796838760376, "perplexity": 7.478825335209925, "muon_lr": 0.00849598445143091, "adamw_lr": 0.00025487953354292724, "tokens_per_sec": 149695.1403231926, "vram_gb": 27.299458026885986, "grad_norm": 0.1455078125, "tokens_processed": 2731540480, "tokens_processed_billions": 2.73154048}
+{"timestamp": 1787544886.452923, "elapsed_hours": 5.072934560047256, "session_id": 2, "event": "train_step", "step": 41690, "epoch": 1, "loss": 2.0481587052345276, "loss_raw": 1.043001651763916, "perplexity": 7.7536112705981814, "muon_lr": 0.008495414219483414, "adamw_lr": 0.0002548624265845024, "tokens_per_sec": 149698.48342099876, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 2732195840, "tokens_processed_billions": 2.73219584}
+{"timestamp": 1787544890.522095, "elapsed_hours": 5.074064885576566, "session_id": 2, "event": "train_step", "step": 41700, "epoch": 1, "loss": 1.9416975820064544, "loss_raw": 2.1001620292663574, "perplexity": 6.97057405355954, "muon_lr": 0.008494843901155924, "adamw_lr": 0.00025484531703467766, "tokens_per_sec": 149701.01486989853, "vram_gb": 27.710740089416504, "grad_norm": 0.134765625, "tokens_processed": 2732851200, "tokens_processed_billions": 2.7328512}
+{"timestamp": 1787544890.5254753, "elapsed_hours": 5.074065824614631, "session_id": 2, "event": "lambda_stats", "step": 41700, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1896701388888888, "lambda_std": 0.2043699421791281, "lambda_all": [2.015625, 1.2421875, 1.203125, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787544894.549728, "elapsed_hours": 5.075183672507604, "session_id": 2, "event": "train_step", "step": 41710, "epoch": 1, "loss": 1.8995511186122895, "loss_raw": 2.1536762714385986, "perplexity": 6.682893942188637, "muon_lr": 0.008494273496464887, "adamw_lr": 0.00025482820489394656, "tokens_per_sec": 149703.88588107648, "vram_gb": 27.710740089416504, "grad_norm": 0.15234375, "tokens_processed": 2733506560, "tokens_processed_billions": 2.73350656}
+{"timestamp": 1787544898.6907873, "elapsed_hours": 5.076333966652553, "session_id": 2, "event": "train_step", "step": 41720, "epoch": 1, "loss": 1.9242927062511443, "loss_raw": 2.484147071838379, "perplexity": 6.850301776389697, "muon_lr": 0.008493132428057965, "adamw_lr": 0.0002547939728417389, "tokens_per_sec": 149705.82549411673, "vram_gb": 27.299458026885986, "grad_norm": 0.1435546875, "tokens_processed": 2734161920, "tokens_processed_billions": 2.73416192}
+{"timestamp": 1787544902.6740577, "elapsed_hours": 5.077440430786875, "session_id": 2, "event": "train_step", "step": 41730, "epoch": 1, "loss": 1.821622896194458, "loss_raw": 1.2235386371612549, "perplexity": 6.181882867569073, "muon_lr": 0.008492561764374986, "adamw_lr": 0.0002547768529312495, "tokens_per_sec": 149709.05785967695, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 2734817280, "tokens_processed_billions": 2.73481728}
+{"timestamp": 1787544906.664389, "elapsed_hours": 5.078548856178919, "session_id": 2, "event": "train_step", "step": 41740, "epoch": 1, "loss": 1.796019537448883, "loss_raw": 1.5085011720657349, "perplexity": 6.025614931263606, "muon_lr": 0.008491991014394268, "adamw_lr": 0.000254759730431828, "tokens_per_sec": 149712.23054179587, "vram_gb": 27.710740089416504, "grad_norm": 0.1328125, "tokens_processed": 2735472640, "tokens_processed_billions": 2.73547264}
+{"timestamp": 1787544910.6747582, "elapsed_hours": 5.079662847386466, "session_id": 2, "event": "train_step", "step": 41750, "epoch": 1, "loss": 1.7715192413330079, "loss_raw": 2.020442485809326, "perplexity": 5.879779383142999, "muon_lr": 0.008491420178132272, "adamw_lr": 0.0002547426053439681, "tokens_per_sec": 149715.23715838796, "vram_gb": 27.710740089416504, "grad_norm": 0.1328125, "tokens_processed": 2736128000, "tokens_processed_billions": 2.736128}
+{"timestamp": 1787544914.873903, "elapsed_hours": 5.080829276707437, "session_id": 2, "event": "train_step", "step": 41760, "epoch": 1, "loss": 1.7636263608932494, "loss_raw": 1.1438910961151123, "perplexity": 5.833553654513285, "muon_lr": 0.008490278246830293, "adamw_lr": 0.00025470834740490874, "tokens_per_sec": 149716.69684172407, "vram_gb": 27.299458026885986, "grad_norm": 0.14453125, "tokens_processed": 2736783360, "tokens_processed_billions": 2.73678336}
+{"timestamp": 1787544918.9232051, "elapsed_hours": 5.081954082714187, "session_id": 2, "event": "train_step", "step": 41770, "epoch": 1, "loss": 1.8373254752159118, "loss_raw": 2.874222755432129, "perplexity": 6.279720513160217, "muon_lr": 0.00848970715182324, "adamw_lr": 0.0002546912145546972, "tokens_per_sec": 149719.38328192977, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2737438720, "tokens_processed_billions": 2.73743872}
+{"timestamp": 1787544922.9622095, "elapsed_hours": 5.083076028492716, "session_id": 2, "event": "train_step", "step": 41780, "epoch": 1, "loss": 1.899420541524887, "loss_raw": 1.2058329582214355, "perplexity": 6.682021366332694, "muon_lr": 0.008489135970600768, "adamw_lr": 0.000254674079118023, "tokens_per_sec": 149722.15277961167, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 2738094080, "tokens_processed_billions": 2.73809408}
+{"timestamp": 1787544926.942806, "elapsed_hours": 5.084181749622027, "session_id": 2, "event": "train_step", "step": 41790, "epoch": 1, "loss": 1.9613271605968476, "loss_raw": 2.5322136878967285, "perplexity": 7.108755267286958, "muon_lr": 0.008488564703179352, "adamw_lr": 0.00025465694109538047, "tokens_per_sec": 149725.3979617296, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2738749440, "tokens_processed_billions": 2.73874944}
+{"timestamp": 1787544931.172756, "elapsed_hours": 5.085356735785802, "session_id": 2, "event": "train_step", "step": 41800, "epoch": 1, "loss": 1.9825715267658233, "loss_raw": 1.2958166599273682, "perplexity": 7.261391860369513, "muon_lr": 0.008487421909805575, "adamw_lr": 0.00025462265729416715, "tokens_per_sec": 149726.60188639822, "vram_gb": 27.299458026885986, "grad_norm": 0.142578125, "tokens_processed": 2739404800, "tokens_processed_billions": 2.7394048}
+{"timestamp": 1787544931.1760366, "elapsed_hours": 5.085357647339503, "session_id": 2, "event": "lambda_stats", "step": 41800, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1896701388888888, "lambda_std": 0.2043699421791281, "lambda_all": [2.015625, 1.2421875, 1.203125, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787544935.1666324, "elapsed_hours": 5.086466145846579, "session_id": 2, "event": "train_step", "step": 41810, "epoch": 1, "loss": 1.97971005320549, "loss_raw": 1.2607777118682861, "perplexity": 7.240643279465184, "muon_lr": 0.00848685038388617, "adamw_lr": 0.0002546055115165851, "tokens_per_sec": 149729.73765905443, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 2740060160, "tokens_processed_billions": 2.74006016}
+{"timestamp": 1787544939.1354322, "elapsed_hours": 5.087568590309885, "session_id": 2, "event": "train_step", "step": 41820, "epoch": 1, "loss": 1.875309578180313, "loss_raw": 0.883042573928833, "perplexity": 6.522838136153032, "muon_lr": 0.00848627877183373, "adamw_lr": 0.00025458836315501185, "tokens_per_sec": 149733.0766777195, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2740715520, "tokens_processed_billions": 2.74071552}
+{"timestamp": 1787544943.107962, "elapsed_hours": 5.088672070768144, "session_id": 2, "event": "train_step", "step": 41830, "epoch": 1, "loss": 1.8882748091220856, "loss_raw": 1.9264663457870483, "perplexity": 6.6079588514738425, "muon_lr": 0.00848570707366474, "adamw_lr": 0.0002545712122099421, "tokens_per_sec": 149736.38300637528, "vram_gb": 27.710740089416504, "grad_norm": 0.1328125, "tokens_processed": 2741370880, "tokens_processed_billions": 2.74137088}
+{"timestamp": 1787544947.285685, "elapsed_hours": 5.089832549426291, "session_id": 2, "event": "train_step", "step": 41840, "epoch": 1, "loss": 1.950575007200241, "loss_raw": 2.7548279762268066, "perplexity": 7.032730288738136, "muon_lr": 0.008484563419043044, "adamw_lr": 0.00025453690257129123, "tokens_per_sec": 149738.0101214053, "vram_gb": 27.299458026885986, "grad_norm": 0.15234375, "tokens_processed": 2742026240, "tokens_processed_billions": 2.74202624}
+{"timestamp": 1787544951.2611516, "elapsed_hours": 5.090936845739683, "session_id": 2, "event": "train_step", "step": 41850, "epoch": 1, "loss": 1.9898264992237091, "loss_raw": 2.5766713619232178, "perplexity": 7.314264621624455, "muon_lr": 0.008483991462623321, "adamw_lr": 0.0002545197438786996, "tokens_per_sec": 149741.2912517297, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 2742681600, "tokens_processed_billions": 2.7426816}
+{"timestamp": 1787544955.234037, "elapsed_hours": 5.0920404251416524, "session_id": 2, "event": "train_step", "step": 41860, "epoch": 1, "loss": 2.0496481239795683, "loss_raw": 0.8228697776794434, "perplexity": 7.765168249020048, "muon_lr": 0.008483419420153007, "adamw_lr": 0.00025450258260459015, "tokens_per_sec": 149744.5921179518, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 2743336960, "tokens_processed_billions": 2.74333696}
+{"timestamp": 1787544959.331297, "elapsed_hours": 5.093178552786509, "session_id": 2, "event": "train_step", "step": 41870, "epoch": 1, "loss": 2.1316496551036836, "loss_raw": 2.4047281742095947, "perplexity": 8.428759895670417, "muon_lr": 0.008482847291648598, "adamw_lr": 0.0002544854187494579, "tokens_per_sec": 149746.87289026595, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 2743992320, "tokens_processed_billions": 2.74399232}
+{"timestamp": 1787544963.4835358, "elapsed_hours": 5.094331952465905, "session_id": 2, "event": "train_step", "step": 41880, "epoch": 1, "loss": 2.0716868841648104, "loss_raw": 1.7015950679779053, "perplexity": 7.9382026581705505, "muon_lr": 0.008481702776603485, "adamw_lr": 0.0002544510832981045, "tokens_per_sec": 149748.70397799465, "vram_gb": 27.299458026885986, "grad_norm": 0.1376953125, "tokens_processed": 2744647680, "tokens_processed_billions": 2.74464768}
+{"timestamp": 1787544967.4928606, "elapsed_hours": 5.095445653663742, "session_id": 2, "event": "train_step", "step": 41890, "epoch": 1, "loss": 2.0821540868282318, "loss_raw": 1.5867284536361694, "perplexity": 8.02172981942147, "muon_lr": 0.008481130390095789, "adamw_lr": 0.00025443391170287366, "tokens_per_sec": 149751.7030327581, "vram_gb": 27.710740089416504, "grad_norm": 0.1630859375, "tokens_processed": 2745303040, "tokens_processed_billions": 2.74530304}
+{"timestamp": 1787544971.563505, "elapsed_hours": 5.09657638821337, "session_id": 2, "event": "train_step", "step": 41900, "epoch": 1, "loss": 2.0541469049453736, "loss_raw": 2.5268325805664062, "perplexity": 7.800180737842608, "muon_lr": 0.008480557917620007, "adamw_lr": 0.0002544167375286002, "tokens_per_sec": 149754.19992684614, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2745958400, "tokens_processed_billions": 2.7459584}
+{"timestamp": 1787544971.5710843, "elapsed_hours": 5.0965784938467875, "session_id": 2, "event": "lambda_stats", "step": 41900, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1896701388888888, "lambda_std": 0.2043699421791281, "lambda_all": [2.015625, 1.2421875, 1.203125, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787544975.5529256, "elapsed_hours": 5.097684560616811, "session_id": 2, "event": "train_step", "step": 41910, "epoch": 1, "loss": 2.0320362758636477, "loss_raw": 1.4502737522125244, "perplexity": 7.6296065353738225, "muon_lr": 0.008479985359192647, "adamw_lr": 0.00025439956077577937, "tokens_per_sec": 149757.35764706112, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2746613760, "tokens_processed_billions": 2.74661376}
+{"timestamp": 1787544979.7473016, "elapsed_hours": 5.098849665058983, "session_id": 2, "event": "train_step", "step": 41920, "epoch": 1, "loss": 2.028276228904724, "loss_raw": 2.5790319442749023, "perplexity": 7.600972722498502, "muon_lr": 0.008478839984549241, "adamw_lr": 0.0002543651995364772, "tokens_per_sec": 149758.84165616691, "vram_gb": 27.299458026885986, "grad_norm": 0.1748046875, "tokens_processed": 2747269120, "tokens_processed_billions": 2.74726912}
+{"timestamp": 1787544983.721759, "elapsed_hours": 5.099953680965635, "session_id": 2, "event": "train_step", "step": 41930, "epoch": 1, "loss": 2.0664974319934846, "loss_raw": 2.7952730655670166, "perplexity": 7.897114440031287, "muon_lr": 0.008478267168366225, "adamw_lr": 0.00025434801505098666, "tokens_per_sec": 149762.11987086592, "vram_gb": 27.710740089416504, "grad_norm": 0.15234375, "tokens_processed": 2747924480, "tokens_processed_billions": 2.74792448}
+{"timestamp": 1787544987.76648, "elapsed_hours": 5.101077214611901, "session_id": 2, "event": "train_step", "step": 41940, "epoch": 1, "loss": 2.006763309240341, "loss_raw": 1.739839792251587, "perplexity": 7.4391999483607245, "muon_lr": 0.00847769426629769, "adamw_lr": 0.0002543308279889306, "tokens_per_sec": 149764.82301595266, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 2748579840, "tokens_processed_billions": 2.74857984}
+{"timestamp": 1787544991.7510355, "elapsed_hours": 5.102184035579364, "session_id": 2, "event": "train_step", "step": 41950, "epoch": 1, "loss": 1.9309519064426421, "loss_raw": 2.5850119590759277, "perplexity": 6.89607153313931, "muon_lr": 0.008477121278360157, "adamw_lr": 0.00025431363835080466, "tokens_per_sec": 149768.0162736177, "vram_gb": 27.710740089416504, "grad_norm": 0.16015625, "tokens_processed": 2749235200, "tokens_processed_billions": 2.7492352}
+{"timestamp": 1787544995.8768682, "elapsed_hours": 5.103330100311173, "session_id": 2, "event": "train_step", "step": 41960, "epoch": 1, "loss": 1.839548455476761, "loss_raw": 1.093348741531372, "perplexity": 6.293695735471246, "muon_lr": 0.008475975044944194, "adamw_lr": 0.0002542792513483258, "tokens_per_sec": 149770.05532586528, "vram_gb": 27.299458026885986, "grad_norm": 0.1552734375, "tokens_processed": 2749890560, "tokens_processed_billions": 2.74989056}
+{"timestamp": 1787544999.9165182, "elapsed_hours": 5.104452225234773, "session_id": 2, "event": "train_step", "step": 41970, "epoch": 1, "loss": 1.7734547555446625, "loss_raw": 1.4819049835205078, "perplexity": 5.89117080026911, "muon_lr": 0.008475401799498822, "adamw_lr": 0.00025426205398496456, "tokens_per_sec": 149772.79719768403, "vram_gb": 27.710740089416504, "grad_norm": 0.1533203125, "tokens_processed": 2750545920, "tokens_processed_billions": 2.75054592}
+{"timestamp": 1787545003.976815, "elapsed_hours": 5.105580085449748, "session_id": 2, "event": "train_step", "step": 41980, "epoch": 1, "loss": 1.831214759349823, "loss_raw": 3.2023844718933105, "perplexity": 6.241463931979006, "muon_lr": 0.00847482846825056, "adamw_lr": 0.0002542448540475167, "tokens_per_sec": 149775.3680696252, "vram_gb": 27.710740089416504, "grad_norm": 0.1337890625, "tokens_processed": 2751201280, "tokens_processed_billions": 2.75120128}
+{"timestamp": 1787545007.9701543, "elapsed_hours": 5.106689346498913, "session_id": 2, "event": "train_step", "step": 41990, "epoch": 1, "loss": 1.8269057416915893, "loss_raw": 2.733144998550415, "perplexity": 6.2146272151507445, "muon_lr": 0.008474255051215941, "adamw_lr": 0.0002542276515364782, "tokens_per_sec": 149778.48513279294, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 2751856640, "tokens_processed_billions": 2.75185664}
+{"timestamp": 1787545012.0882928, "elapsed_hours": 5.107833273741934, "session_id": 2, "event": "train_step", "step": 42000, "epoch": 1, "loss": 1.8908304572105408, "loss_raw": 2.0165677070617676, "perplexity": 6.624868066679647, "muon_lr": 0.008473107959853786, "adamw_lr": 0.00025419323879561354, "tokens_per_sec": 149780.58259890942, "vram_gb": 27.299458026885986, "grad_norm": 0.146484375, "tokens_processed": 2752512000, "tokens_processed_billions": 2.752512}
+{"timestamp": 1787545012.091389, "elapsed_hours": 5.107834133903186, "session_id": 2, "event": "lambda_stats", "step": 42000, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1896701388888888, "lambda_std": 0.2043699421791281, "lambda_all": [2.015625, 1.2421875, 1.203125, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787545025.6599624, "elapsed_hours": 5.11160318215688, "session_id": 2, "event": "validation", "step": 42000, "epoch": 1, "val_loss": 1.967289936542511, "val_perplexity": 7.151269808574514, "is_best": true, "best_loss": 1.967289936542511, "best_perplexity": 7.151269808574514, "train_loss": 2.0165677070617676, "train_loss_avg": 1.8908304572105408}