Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -8077,3 +8077,59 @@
 {"timestamp": 1787557794.4671593, "elapsed_hours": 8.6584940699074, "session_id": 2, "event": "train_step", "step": 71500, "epoch": 1, "loss": 1.8370429503917693, "loss_raw": 1.2519603967666626, "perplexity": 6.277946586827484, "muon_lr": 0.005929668340480228, "adamw_lr": 0.0001778900502144068, "tokens_per_sec": 150382.64429842896, "vram_gb": 27.710740089416504, "grad_norm": 0.138671875, "tokens_processed": 4685824000, "tokens_processed_billions": 4.685824}
 {"timestamp": 1787557794.4704583, "elapsed_hours": 8.658494986626836, "session_id": 2, "event": "lambda_stats", "step": 71500, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 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": 1787557808.0724983, "elapsed_hours": 8.662273331218296, "session_id": 2, "event": "validation", "step": 71500, "epoch": 1, "val_loss": 1.8583039635419845, "val_perplexity": 6.412851113513861, "is_best": true, "best_loss": 1.8583039635419845, "best_perplexity": 6.412851113513861, "train_loss": 1.2519603967666626, "train_loss_avg": 1.8370429503917693}
+{"timestamp": 1787557812.7983882, "elapsed_hours": 8.663586078153717, "session_id": 2, "event": "train_step", "step": 71510, "epoch": 1, "loss": 1.8255861282348633, "loss_raw": 1.4020487070083618, "perplexity": 6.206431718082494, "muon_lr": 0.0059289076544702946, "adamw_lr": 0.00017786722963410878, "tokens_per_sec": 150315.2750872623, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4686479360, "tokens_processed_billions": 4.68647936}
+{"timestamp": 1787557817.0349474, "elapsed_hours": 8.664762900074322, "session_id": 2, "event": "train_step", "step": 71520, "epoch": 1, "loss": 1.7100942170619964, "loss_raw": 2.2443864345550537, "perplexity": 5.529482424670982, "muon_lr": 0.005927386245366605, "adamw_lr": 0.00017782158736099813, "tokens_per_sec": 150315.83996490404, "vram_gb": 27.299458026885986, "grad_norm": 0.13671875, "tokens_processed": 4687134720, "tokens_processed_billions": 4.68713472}
+{"timestamp": 1787557821.0108392, "elapsed_hours": 8.66586731447114, "session_id": 2, "event": "train_step", "step": 71530, "epoch": 1, "loss": 1.6202877652645111, "loss_raw": 0.6592538952827454, "perplexity": 5.054544629675775, "muon_lr": 0.0059266255223167225, "adamw_lr": 0.00017779876566950163, "tokens_per_sec": 150317.69082802517, "vram_gb": 27.710740089416504, "grad_norm": 0.125, "tokens_processed": 4687790080, "tokens_processed_billions": 4.68779008}
+{"timestamp": 1787557825.005641, "elapsed_hours": 8.666976981560389, "session_id": 2, "event": "train_step", "step": 71540, "epoch": 1, "loss": 1.6121530795097352, "loss_raw": 1.6079039573669434, "perplexity": 5.0135942824006, "muon_lr": 0.005925864786964062, "adamw_lr": 0.00017777594360892182, "tokens_per_sec": 150319.45010014423, "vram_gb": 27.710740089416504, "grad_norm": 0.12255859375, "tokens_processed": 4688445440, "tokens_processed_billions": 4.68844544}
+{"timestamp": 1787557829.0055044, "elapsed_hours": 8.668088054921892, "session_id": 2, "event": "train_step", "step": 71550, "epoch": 1, "loss": 1.6463576173782348, "loss_raw": 1.847474455833435, "perplexity": 5.188048512820726, "muon_lr": 0.005925104039330564, "adamw_lr": 0.00017775312117991685, "tokens_per_sec": 150321.1847586637, "vram_gb": 27.710740089416504, "grad_norm": 0.1572265625, "tokens_processed": 4689100800, "tokens_processed_billions": 4.6891008}
+{"timestamp": 1787557833.1348078, "elapsed_hours": 8.669235083593263, "session_id": 2, "event": "train_step", "step": 71560, "epoch": 1, "loss": 1.6836410772800445, "loss_raw": 1.6205053329467773, "perplexity": 5.385127985047948, "muon_lr": 0.0059235825073088045, "adamw_lr": 0.00017770747521926407, "tokens_per_sec": 150322.2950332602, "vram_gb": 27.299458026885986, "grad_norm": 0.1474609375, "tokens_processed": 4689756160, "tokens_processed_billions": 4.68975616}
+{"timestamp": 1787557837.1410296, "elapsed_hours": 8.670347922907935, "session_id": 2, "event": "train_step", "step": 71570, "epoch": 1, "loss": 1.6591002142429352, "loss_raw": 0.7677602171897888, "perplexity": 5.254580719825602, "muon_lr": 0.00592282172296442, "adamw_lr": 0.00017768465168893256, "tokens_per_sec": 150323.99870369033, "vram_gb": 27.710740089416504, "grad_norm": 0.14453125, "tokens_processed": 4690411520, "tokens_processed_billions": 4.69041152}
+{"timestamp": 1787557841.1141636, "elapsed_hours": 8.67145157125261, "session_id": 2, "event": "train_step", "step": 71580, "epoch": 1, "loss": 1.7709195625782013, "loss_raw": 1.5680230855941772, "perplexity": 5.876254461379959, "muon_lr": 0.005922060926426952, "adamw_lr": 0.00017766182779280852, "tokens_per_sec": 150325.85978929378, "vram_gb": 27.710740089416504, "grad_norm": 0.13671875, "tokens_processed": 4691066880, "tokens_processed_billions": 4.69106688}
+{"timestamp": 1787557845.0858114, "elapsed_hours": 8.672554806802008, "session_id": 2, "event": "train_step", "step": 71590, "epoch": 1, "loss": 1.8139792549610139, "loss_raw": 1.6101646423339844, "perplexity": 6.134810902570893, "muon_lr": 0.005921300117718338, "adamw_lr": 0.00017763900353155008, "tokens_per_sec": 150327.72841673077, "vram_gb": 27.710740089416504, "grad_norm": 0.12890625, "tokens_processed": 4691722240, "tokens_processed_billions": 4.69172224}
+{"timestamp": 1787557849.2244518, "elapsed_hours": 8.673704429003928, "session_id": 2, "event": "train_step", "step": 71600, "epoch": 1, "loss": 1.8577310788631438, "loss_raw": 2.2188944816589355, "perplexity": 6.409178341501099, "muon_lr": 0.005919778463875434, "adamw_lr": 0.00017759335391626297, "tokens_per_sec": 150328.79199937303, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 4692377600, "tokens_processed_billions": 4.6923776}
+{"timestamp": 1787557849.2280724, "elapsed_hours": 8.673705434799194, "session_id": 2, "event": "lambda_stats", "step": 71600, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 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": 1787557853.2046375, "elapsed_hours": 8.674810036155913, "session_id": 2, "event": "train_step", "step": 71610, "epoch": 1, "loss": 1.8174384367465972, "loss_raw": 0.956718921661377, "perplexity": 6.15606907544744, "muon_lr": 0.005919017618785027, "adamw_lr": 0.00017757052856355075, "tokens_per_sec": 150330.61885903732, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 4693032960, "tokens_processed_billions": 4.69303296}
+{"timestamp": 1787557857.1643848, "elapsed_hours": 8.675909965965483, "session_id": 2, "event": "train_step", "step": 71620, "epoch": 1, "loss": 1.7298196959495544, "loss_raw": 2.046149253845215, "perplexity": 5.639636967363521, "muon_lr": 0.0059182567616112335, "adamw_lr": 0.00017754770284833697, "tokens_per_sec": 150332.54323470066, "vram_gb": 27.710740089416504, "grad_norm": 0.1103515625, "tokens_processed": 4693688320, "tokens_processed_billions": 4.69368832}
+{"timestamp": 1787557861.239632, "elapsed_hours": 8.677041979167196, "session_id": 2, "event": "train_step", "step": 71630, "epoch": 1, "loss": 1.6943672001361847, "loss_raw": 1.1083087921142578, "perplexity": 5.443200418673428, "muon_lr": 0.005917495892375999, "adamw_lr": 0.00017752487677127993, "tokens_per_sec": 150333.91135159583, "vram_gb": 27.710740089416504, "grad_norm": 0.1533203125, "tokens_processed": 4694343680, "tokens_processed_billions": 4.69434368}
+{"timestamp": 1787557865.3713179, "elapsed_hours": 8.67818966958258, "session_id": 2, "event": "train_step", "step": 71640, "epoch": 1, "loss": 1.6444635665416718, "loss_raw": 1.246086835861206, "perplexity": 5.178231385199133, "muon_lr": 0.0059159741178089675, "adamw_lr": 0.00017747922353426898, "tokens_per_sec": 150335.00689365255, "vram_gb": 27.299458026885986, "grad_norm": 0.130859375, "tokens_processed": 4694999040, "tokens_processed_billions": 4.69499904}
+{"timestamp": 1787557869.4147084, "elapsed_hours": 8.679312833746275, "session_id": 2, "event": "train_step", "step": 71650, "epoch": 1, "loss": 1.6160682225227356, "loss_raw": 1.4375741481781006, "perplexity": 5.033261696271959, "muon_lr": 0.0059152132125210545, "adamw_lr": 0.00017745639637563158, "tokens_per_sec": 150336.52825657977, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 4695654400, "tokens_processed_billions": 4.6956544}
+{"timestamp": 1787557873.4224267, "elapsed_hours": 8.680426088704003, "session_id": 2, "event": "train_step", "step": 71660, "epoch": 1, "loss": 1.6909208846092225, "loss_raw": 1.9005988836288452, "perplexity": 5.424473720145028, "muon_lr": 0.005914452295259468, "adamw_lr": 0.00017743356885778396, "tokens_per_sec": 150338.21994197922, "vram_gb": 27.710740089416504, "grad_norm": 0.1474609375, "tokens_processed": 4696309760, "tokens_processed_billions": 4.69630976}
+{"timestamp": 1787557877.3988142, "elapsed_hours": 8.681530640787548, "session_id": 2, "event": "train_step", "step": 71670, "epoch": 1, "loss": 1.8862854206562043, "loss_raw": 2.7198262214660645, "perplexity": 6.594826121733484, "muon_lr": 0.005913691366046148, "adamw_lr": 0.0001774107409813844, "tokens_per_sec": 150340.06237929442, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 4696965120, "tokens_processed_billions": 4.69696512}
+{"timestamp": 1787557881.5956426, "elapsed_hours": 8.682696426444583, "session_id": 2, "event": "train_step", "step": 71680, "epoch": 1, "loss": 1.9225231897830963, "loss_raw": 1.8578855991363525, "perplexity": 6.838190773055352, "muon_lr": 0.005912169471852088, "adamw_lr": 0.00017736508415556258, "tokens_per_sec": 150340.84309874466, "vram_gb": 27.299458026885986, "grad_norm": 0.287109375, "tokens_processed": 4697620480, "tokens_processed_billions": 4.69762048}
+{"timestamp": 1787557885.5531938, "elapsed_hours": 8.683795746167501, "session_id": 2, "event": "train_step", "step": 71690, "epoch": 1, "loss": 1.95415846824646, "loss_raw": 1.7044175863265991, "perplexity": 7.057977012084723, "muon_lr": 0.005911408506915233, "adamw_lr": 0.00017734225520745696, "tokens_per_sec": 150342.77558753366, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 4698275840, "tokens_processed_billions": 4.69827584}
+{"timestamp": 1787557889.5199847, "elapsed_hours": 8.68489763908916, "session_id": 2, "event": "train_step", "step": 71700, "epoch": 1, "loss": 1.884490602016449, "loss_raw": 2.2922613620758057, "perplexity": 6.5830002207326626, "muon_lr": 0.005910647530114424, "adamw_lr": 0.00017731942590343267, "tokens_per_sec": 150344.6624274721, "vram_gb": 27.710740089416504, "grad_norm": 0.14453125, "tokens_processed": 4698931200, "tokens_processed_billions": 4.6989312}
+{"timestamp": 1787557889.5259526, "elapsed_hours": 8.684899290468957, "session_id": 2, "event": "lambda_stats", "step": 71700, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 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": 1787557893.5003364, "elapsed_hours": 8.686003285911347, "session_id": 2, "event": "train_step", "step": 71710, "epoch": 1, "loss": 1.8467391180992125, "loss_raw": 1.4043002128601074, "perplexity": 6.339114678600655, "muon_lr": 0.005909886541471601, "adamw_lr": 0.00017729659624414796, "tokens_per_sec": 150346.48430240326, "vram_gb": 27.710740089416504, "grad_norm": 0.138671875, "tokens_processed": 4699586560, "tokens_processed_billions": 4.69958656}
+{"timestamp": 1787557897.63347, "elapsed_hours": 8.687151378591855, "session_id": 2, "event": "train_step", "step": 71720, "epoch": 1, "loss": 1.8001812541484832, "loss_raw": 2.784717321395874, "perplexity": 6.050744087493471, "muon_lr": 0.0059083645287476986, "adamw_lr": 0.00017725093586243093, "tokens_per_sec": 150347.5703687054, "vram_gb": 27.299458026885986, "grad_norm": 0.138671875, "tokens_processed": 4700241920, "tokens_processed_billions": 4.70024192}
+{"timestamp": 1787557901.6050663, "elapsed_hours": 8.688254599902365, "session_id": 2, "event": "train_step", "step": 71730, "epoch": 1, "loss": 1.7752120900154114, "loss_raw": 0.7929208874702454, "perplexity": 5.90153265974927, "muon_lr": 0.005907603504710511, "adamw_lr": 0.00017722810514131528, "tokens_per_sec": 150349.43307946948, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 4700897280, "tokens_processed_billions": 4.70089728}
+{"timestamp": 1787557905.571872, "elapsed_hours": 8.689356490241156, "session_id": 2, "event": "train_step", "step": 71740, "epoch": 1, "loss": 1.7900162363052368, "loss_raw": 1.1634531021118164, "perplexity": 5.989549713751026, "muon_lr": 0.0059068424689190935, "adamw_lr": 0.00017720527406757275, "tokens_per_sec": 150351.31806044857, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 4701552640, "tokens_processed_billions": 4.70155264}
+{"timestamp": 1787557909.5405946, "elapsed_hours": 8.690458913180564, "session_id": 2, "event": "train_step", "step": 71750, "epoch": 1, "loss": 1.8277226722240447, "loss_raw": 2.1207375526428223, "perplexity": 6.219706208180404, "muon_lr": 0.0059060814213953905, "adamw_lr": 0.00017718244264186165, "tokens_per_sec": 150353.19367210465, "vram_gb": 27.710740089416504, "grad_norm": 0.134765625, "tokens_processed": 4702208000, "tokens_processed_billions": 4.702208}
+{"timestamp": 1787557913.6604774, "elapsed_hours": 8.691603325274256, "session_id": 2, "event": "train_step", "step": 71760, "epoch": 1, "loss": 1.7865689146518706, "loss_raw": 1.5758947134017944, "perplexity": 5.968937358451418, "muon_lr": 0.0059045592912389195, "adamw_lr": 0.00017713677873716752, "tokens_per_sec": 150354.3419496466, "vram_gb": 27.299458026885986, "grad_norm": 0.134765625, "tokens_processed": 4702863360, "tokens_processed_billions": 4.70286336}
+{"timestamp": 1787557917.6318283, "elapsed_hours": 8.692706478039424, "session_id": 2, "event": "train_step", "step": 71770, "epoch": 1, "loss": 1.7198425710201264, "loss_raw": 1.9906493425369263, "perplexity": 5.583649366856369, "muon_lr": 0.005903798208650046, "adamw_lr": 0.00017711394625950132, "tokens_per_sec": 150356.20431316402, "vram_gb": 27.710740089416504, "grad_norm": 0.1318359375, "tokens_processed": 4703518720, "tokens_processed_billions": 4.70351872}
+{"timestamp": 1787557921.5992794, "elapsed_hours": 8.693808547920652, "session_id": 2, "event": "train_step", "step": 71780, "epoch": 1, "loss": 1.7153715443611146, "loss_raw": 1.7997117042541504, "perplexity": 5.558740447398311, "muon_lr": 0.005903037114416677, "adamw_lr": 0.00017709111343250028, "tokens_per_sec": 150358.08480143046, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4704174080, "tokens_processed_billions": 4.70417408}
+{"timestamp": 1787557925.5837998, "elapsed_hours": 8.694915359020234, "session_id": 2, "event": "train_step", "step": 71790, "epoch": 1, "loss": 1.7371517753601073, "loss_raw": 0.863479495048523, "perplexity": 5.681139196346525, "muon_lr": 0.005902276008560762, "adamw_lr": 0.0001770682802568228, "tokens_per_sec": 150359.88255928888, "vram_gb": 27.710740089416504, "grad_norm": 0.166015625, "tokens_processed": 4704829440, "tokens_processed_billions": 4.70482944}
+{"timestamp": 1787557929.7160547, "elapsed_hours": 8.696063207785288, "session_id": 2, "event": "train_step", "step": 71800, "epoch": 1, "loss": 1.7448239123821259, "loss_raw": 1.7459907531738281, "perplexity": 5.724893303836188, "muon_lr": 0.0059007537620690805, "adamw_lr": 0.00017702261286207236, "tokens_per_sec": 150360.9699283616, "vram_gb": 27.299458026885986, "grad_norm": 0.126953125, "tokens_processed": 4705484800, "tokens_processed_billions": 4.7054848}
+{"timestamp": 1787557929.7194545, "elapsed_hours": 8.696064152121544, "session_id": 2, "event": "lambda_stats", "step": 71800, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 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": 1787557933.6915164, "elapsed_hours": 8.69716750257545, "session_id": 2, "event": "train_step", "step": 71810, "epoch": 1, "loss": 1.8629793894290925, "loss_raw": 2.541487216949463, "perplexity": 6.442904124186631, "muon_lr": 0.005899992621477213, "adamw_lr": 0.00017699977864431635, "tokens_per_sec": 150362.8105965507, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 4706140160, "tokens_processed_billions": 4.70614016}
+{"timestamp": 1787557937.6603765, "elapsed_hours": 8.698269963728057, "session_id": 2, "event": "train_step", "step": 71820, "epoch": 1, "loss": 1.9198631131649018, "loss_raw": 1.054396629333496, "perplexity": 6.820024833776215, "muon_lr": 0.0058992314693505936, "adamw_lr": 0.00017697694408051775, "tokens_per_sec": 150364.68254741747, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 4706795520, "tokens_processed_billions": 4.70679552}
+{"timestamp": 1787557941.6377935, "elapsed_hours": 8.699374801913898, "session_id": 2, "event": "train_step", "step": 71830, "epoch": 1, "loss": 1.958553248643875, "loss_raw": 2.470518112182617, "perplexity": 7.089063530279252, "muon_lr": 0.005898470305711173, "adamw_lr": 0.00017695410917133514, "tokens_per_sec": 150366.51270260845, "vram_gb": 27.710740089416504, "grad_norm": 0.13671875, "tokens_processed": 4707450880, "tokens_processed_billions": 4.70745088}
+{"timestamp": 1787557945.7542102, "elapsed_hours": 8.700518250862757, "session_id": 2, "event": "train_step", "step": 71840, "epoch": 1, "loss": 1.9899924409389496, "loss_raw": 2.890744686126709, "perplexity": 7.3154784639522425, "muon_lr": 0.0058969479439817205, "adamw_lr": 0.00017690843831945157, "tokens_per_sec": 150367.6743883832, "vram_gb": 27.299458026885986, "grad_norm": 0.1435546875, "tokens_processed": 4708106240, "tokens_processed_billions": 4.70810624}
+{"timestamp": 1787557949.7817147, "elapsed_hours": 8.701637002097236, "session_id": 2, "event": "train_step", "step": 71850, "epoch": 1, "loss": 2.0486348271369934, "loss_raw": 1.8472620248794556, "perplexity": 7.757303813727893, "muon_lr": 0.005896186745935592, "adamw_lr": 0.00017688560237806773, "tokens_per_sec": 150369.26343257536, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 4708761600, "tokens_processed_billions": 4.7087616}
+{"timestamp": 1787557953.7576182, "elapsed_hours": 8.702741419672966, "session_id": 2, "event": "train_step", "step": 71860, "epoch": 1, "loss": 2.10413898229599, "loss_raw": 2.4080796241760254, "perplexity": 8.20003958990699, "muon_lr": 0.005895425536464463, "adamw_lr": 0.00017686276609393383, "tokens_per_sec": 150371.09980821563, "vram_gb": 27.710740089416504, "grad_norm": 0.1611328125, "tokens_processed": 4709416960, "tokens_processed_billions": 4.70941696}
+{"timestamp": 1787557957.8352056, "elapsed_hours": 8.70387408276399, "session_id": 2, "event": "train_step", "step": 71870, "epoch": 1, "loss": 2.070735454559326, "loss_raw": 2.2356343269348145, "perplexity": 7.930653608911174, "muon_lr": 0.005894664315590286, "adamw_lr": 0.0001768399294677085, "tokens_per_sec": 150372.44735944315, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 4710072320, "tokens_processed_billions": 4.71007232}
+{"timestamp": 1787557962.0279484, "elapsed_hours": 8.705038733681043, "session_id": 2, "event": "train_step", "step": 71880, "epoch": 1, "loss": 2.0184382390975952, "loss_raw": 0.8683920502662659, "perplexity": 7.5265610610801374, "muon_lr": 0.005893141839720588, "adamw_lr": 0.0001767942551916176, "tokens_per_sec": 150373.24146347234, "vram_gb": 27.299458026885986, "grad_norm": 0.134765625, "tokens_processed": 4710727680, "tokens_processed_billions": 4.71072768}
+{"timestamp": 1787557965.9969547, "elapsed_hours": 8.706141235298581, "session_id": 2, "event": "train_step", "step": 71890, "epoch": 1, "loss": 1.9809718549251556, "loss_raw": 1.3152586221694946, "perplexity": 7.24978530210369, "muon_lr": 0.005892380584768974, "adamw_lr": 0.00017677141754306915, "tokens_per_sec": 150375.10940556647, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 4711383040, "tokens_processed_billions": 4.71138304}
+{"timestamp": 1787557970.0453382, "elapsed_hours": 8.707265786263678, "session_id": 2, "event": "train_step", "step": 71900, "epoch": 1, "loss": 1.888254680633545, "loss_raw": 0.5960104465484619, "perplexity": 6.607825844588442, "muon_lr": 0.005891619318502117, "adamw_lr": 0.00017674857955506345, "tokens_per_sec": 150376.59633978803, "vram_gb": 27.710740089416504, "grad_norm": 0.12158203125, "tokens_processed": 4712038400, "tokens_processed_billions": 4.7120384}
+{"timestamp": 1787557970.0487318, "elapsed_hours": 8.707266729341613, "session_id": 2, "event": "lambda_stats", "step": 71900, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 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": 1787557974.0229242, "elapsed_hours": 8.70837067147096, "session_id": 2, "event": "train_step", "step": 71910, "epoch": 1, "loss": 1.7037390959262848, "loss_raw": 0.9882558584213257, "perplexity": 5.49445331904412, "muon_lr": 0.005890858040941975, "adamw_lr": 0.00017672574122825922, "tokens_per_sec": 150378.42234919104, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 4712693760, "tokens_processed_billions": 4.71269376}
+{"timestamp": 1787557978.2159977, "elapsed_hours": 8.709535414046712, "session_id": 2, "event": "train_step", "step": 71920, "epoch": 1, "loss": 1.78189381480217, "loss_raw": 3.103494167327881, "perplexity": 5.941097108984749, "muon_lr": 0.005889335452029639, "adamw_lr": 0.00017668006356088913, "tokens_per_sec": 150379.21363320286, "vram_gb": 27.299458026885986, "grad_norm": 0.169921875, "tokens_processed": 4713349120, "tokens_processed_billions": 4.71334912}
+{"timestamp": 1787557982.1768737, "elapsed_hours": 8.710635657376713, "session_id": 2, "event": "train_step", "step": 71930, "epoch": 1, "loss": 1.8659064495563507, "loss_raw": 2.59892201423645, "perplexity": 6.461790519274908, "muon_lr": 0.005888574140721354, "adamw_lr": 0.00017665722422164056, "tokens_per_sec": 150381.1194813594, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 4714004480, "tokens_processed_billions": 4.71400448}
+{"timestamp": 1787557986.1620111, "elapsed_hours": 8.711742640005218, "session_id": 2, "event": "train_step", "step": 71940, "epoch": 1, "loss": 1.8320526576042175, "loss_raw": 3.245433807373047, "perplexity": 6.2466958353077136, "muon_lr": 0.005887812818207595, "adamw_lr": 0.0001766343845462278, "tokens_per_sec": 150382.90798572902, "vram_gb": 27.710740089416504, "grad_norm": 0.1298828125, "tokens_processed": 4714659840, "tokens_processed_billions": 4.71465984}
+{"timestamp": 1787557990.129648, "elapsed_hours": 8.712844761278895, "session_id": 2, "event": "train_step", "step": 71950, "epoch": 1, "loss": 1.874788329601288, "loss_raw": 1.186095952987671, "perplexity": 6.519439002017233, "muon_lr": 0.005887051484510318, "adamw_lr": 0.0001766115445353095, "tokens_per_sec": 150384.77946537017, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 4715315200, "tokens_processed_billions": 4.7153152}
+{"timestamp": 1787557994.2769113, "elapsed_hours": 8.71399677892526, "session_id": 2, "event": "train_step", "step": 71960, "epoch": 1, "loss": 1.916660453081131, "loss_raw": 1.5694671869277954, "perplexity": 6.798217551766398, "muon_lr": 0.0058855287836530315, "adamw_lr": 0.0001765658635095909, "tokens_per_sec": 150385.7895034605, "vram_gb": 27.299458026885986, "grad_norm": 0.1455078125, "tokens_processed": 4715970560, "tokens_processed_billions": 4.71597056}
+{"timestamp": 1787557998.2636974, "elapsed_hours": 8.715104219383663, "session_id": 2, "event": "train_step", "step": 71970, "epoch": 1, "loss": 1.869598196744919, "loss_raw": 1.7715001106262207, "perplexity": 6.485689904357687, "muon_lr": 0.005884767416536931, "adamw_lr": 0.00017654302249610788, "tokens_per_sec": 150387.56908308802, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4716625920, "tokens_processed_billions": 4.71662592}
+{"timestamp": 1787558002.2297437, "elapsed_hours": 8.716205898986923, "session_id": 2, "event": "train_step", "step": 71980, "epoch": 1, "loss": 1.7872326743602753, "loss_raw": 1.261677861213684, "perplexity": 5.972900613751031, "muon_lr": 0.0058840060383251345, "adamw_lr": 0.00017652018114975397, "tokens_per_sec": 150389.4472687769, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 4717281280, "tokens_processed_billions": 4.71728128}
+{"timestamp": 1787558006.2028093, "elapsed_hours": 8.717309528258111, "session_id": 2, "event": "train_step", "step": 71990, "epoch": 1, "loss": 1.8494151103496552, "loss_raw": 2.3671865463256836, "perplexity": 6.356100817607046, "muon_lr": 0.0058832446490395965, "adamw_lr": 0.00017649733947118786, "tokens_per_sec": 150391.2915706612, "vram_gb": 27.710740089416504, "grad_norm": 0.1298828125, "tokens_processed": 4717936640, "tokens_processed_billions": 4.71793664}
+{"timestamp": 1787558010.4431689, "elapsed_hours": 8.718487406041888, "session_id": 2, "event": "train_step", "step": 72000, "epoch": 1, "loss": 1.8465044116973877, "loss_raw": 1.6065235137939453, "perplexity": 6.337627022391732, "muon_lr": 0.005881721837335127, "adamw_lr": 0.00017645165512005377, "tokens_per_sec": 150391.85388872505, "vram_gb": 27.299458026885986, "grad_norm": 0.140625, "tokens_processed": 4718592000, "tokens_processed_billions": 4.718592}
+{"timestamp": 1787558010.4468613, "elapsed_hours": 8.718488431837823, "session_id": 2, "event": "lambda_stats", "step": 72000, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 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": 1787558024.100016, "elapsed_hours": 8.722280974785487, "session_id": 2, "event": "validation", "step": 72000, "epoch": 1, "val_loss": 1.8571415823698043, "val_perplexity": 6.4054012667391405, "is_best": true, "best_loss": 1.8571415823698043, "best_perplexity": 6.4054012667391405, "train_loss": 1.6065235137939453, "train_loss_avg": 1.8465044116973877}