Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -8981,3 +8981,59 @@
 {"timestamp": 1787561348.548435, "elapsed_hours": 9.64573886891206, "session_id": 2, "event": "train_step", "step": 79500, "epoch": 1, "loss": 2.0218741703033447, "loss_raw": 2.2090981006622314, "perplexity": 7.552466285901454, "muon_lr": 0.005166458872626068, "adamw_lr": 0.000154993766178782, "tokens_per_sec": 150089.2668057516, "vram_gb": 27.710740089416504, "grad_norm": 0.13671875, "tokens_processed": 5210112000, "tokens_processed_billions": 5.210112}
 {"timestamp": 1787561348.55185, "elapsed_hours": 9.645739817420642, "session_id": 2, "event": "lambda_stats", "step": 79500, "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": 1787561362.1213796, "elapsed_hours": 9.649509131511053, "session_id": 2, "event": "validation", "step": 79500, "epoch": 1, "val_loss": 1.838739149570465, "val_perplexity": 6.288604270894176, "is_best": true, "best_loss": 1.838739149570465, "best_perplexity": 6.288604270894176, "train_loss": 2.2090981006622314, "train_loss_avg": 2.0218741703033447}
+{"timestamp": 1787561367.4244, "elapsed_hours": 9.650982192423609, "session_id": 2, "event": "train_step", "step": 79510, "epoch": 1, "loss": 1.9089910900592804, "loss_raw": 1.5547780990600586, "perplexity": 6.746278975938383, "muon_lr": 0.005165696808250165, "adamw_lr": 0.0001549709042475049, "tokens_per_sec": 150026.5668206171, "vram_gb": 27.710740089416504, "grad_norm": 0.1572265625, "tokens_processed": 5210767360, "tokens_processed_billions": 5.21076736}
+{"timestamp": 1787561371.6410072, "elapsed_hours": 9.652153472304343, "session_id": 2, "event": "train_step", "step": 79520, "epoch": 1, "loss": 1.8450381350517273, "loss_raw": 0.9139338135719299, "perplexity": 6.328341117416392, "muon_lr": 0.005164172708441648, "adamw_lr": 0.0001549251812532494, "tokens_per_sec": 150027.2214547855, "vram_gb": 27.299458026885986, "grad_norm": 0.1396484375, "tokens_processed": 5211422720, "tokens_processed_billions": 5.21142272}
+{"timestamp": 1787561375.7531848, "elapsed_hours": 9.65329574379656, "session_id": 2, "event": "train_step", "step": 79530, "epoch": 1, "loss": 1.8768560099601745, "loss_raw": 1.4760725498199463, "perplexity": 6.532933063888411, "muon_lr": 0.005163410673052986, "adamw_lr": 0.00015490232019158953, "tokens_per_sec": 150028.3278790759, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 5212078080, "tokens_processed_billions": 5.21207808}
+{"timestamp": 1787561379.843983, "elapsed_hours": 9.654432076613109, "session_id": 2, "event": "train_step", "step": 79540, "epoch": 1, "loss": 1.8423511755466462, "loss_raw": 3.2860159873962402, "perplexity": 6.311359945177284, "muon_lr": 0.0051626486473706915, "adamw_lr": 0.00015487945942112069, "tokens_per_sec": 150029.52579493495, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 5212733440, "tokens_processed_billions": 5.21273344}
+{"timestamp": 1787561383.8283148, "elapsed_hours": 9.65553883559174, "session_id": 2, "event": "train_step", "step": 79550, "epoch": 1, "loss": 1.8265593481063842, "loss_raw": 1.6795895099639893, "perplexity": 6.212474880947878, "muon_lr": 0.005161886631416737, "adamw_lr": 0.00015485659894250205, "tokens_per_sec": 150031.18337339768, "vram_gb": 27.710740089416504, "grad_norm": 0.1357421875, "tokens_processed": 5213388800, "tokens_processed_billions": 5.2133888}
+{"timestamp": 1787561388.0558362, "elapsed_hours": 9.656713147097165, "session_id": 2, "event": "train_step", "step": 79560, "epoch": 1, "loss": 1.8389999043941498, "loss_raw": 0.8504005670547485, "perplexity": 6.290244268601421, "muon_lr": 0.005160362628781744, "adamw_lr": 0.00015481087886345227, "tokens_per_sec": 150031.79019208657, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 5214044160, "tokens_processed_billions": 5.21404416}
+{"timestamp": 1787561392.1142168, "elapsed_hours": 9.65784047484398, "session_id": 2, "event": "train_step", "step": 79570, "epoch": 1, "loss": 1.8418726241588592, "loss_raw": 0.7493006587028503, "perplexity": 6.308340357688215, "muon_lr": 0.005159600642144654, "adamw_lr": 0.00015478801926433958, "tokens_per_sec": 150033.12752373223, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 5214699520, "tokens_processed_billions": 5.21469952}
+{"timestamp": 1787561396.2045348, "elapsed_hours": 9.658976674344805, "session_id": 2, "event": "train_step", "step": 79580, "epoch": 1, "loss": 1.8254765951633454, "loss_raw": 1.962019920349121, "perplexity": 6.2057519457826995, "muon_lr": 0.005158838665323803, "adamw_lr": 0.00015476515995971403, "tokens_per_sec": 150034.32639269903, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 5215354880, "tokens_processed_billions": 5.21535488}
+{"timestamp": 1787561400.3481405, "elapsed_hours": 9.660127676261796, "session_id": 2, "event": "train_step", "step": 79590, "epoch": 1, "loss": 1.8406474316120147, "loss_raw": 1.4172505140304565, "perplexity": 6.300616158880652, "muon_lr": 0.0051580766983411585, "adamw_lr": 0.00015474230095023473, "tokens_per_sec": 150035.29557165963, "vram_gb": 27.710740089416504, "grad_norm": 0.125, "tokens_processed": 5216010240, "tokens_processed_billions": 5.21601024}
+{"timestamp": 1787561404.487514, "elapsed_hours": 9.661277501980464, "session_id": 2, "event": "train_step", "step": 79600, "epoch": 1, "loss": 1.8742331194877624, "loss_raw": 1.4379923343658447, "perplexity": 6.515820348198355, "muon_lr": 0.005156552793978399, "adamw_lr": 0.00015469658381935192, "tokens_per_sec": 150036.28138919512, "vram_gb": 27.299458026885986, "grad_norm": 0.142578125, "tokens_processed": 5216665600, "tokens_processed_billions": 5.2166656}
+{"timestamp": 1787561404.4908128, "elapsed_hours": 9.661278418434991, "session_id": 2, "event": "lambda_stats", "step": 79600, "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": 1787561408.5588474, "elapsed_hours": 9.662408427993457, "session_id": 2, "event": "train_step", "step": 79610, "epoch": 1, "loss": 1.8610530829429626, "loss_raw": 1.088817834854126, "perplexity": 6.430505062213503, "muon_lr": 0.005155790856642224, "adamw_lr": 0.00015467372569926666, "tokens_per_sec": 150037.56187967648, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 5217320960, "tokens_processed_billions": 5.21732096}
+{"timestamp": 1787561412.6525419, "elapsed_hours": 9.663545565274026, "session_id": 2, "event": "train_step", "step": 79620, "epoch": 1, "loss": 1.8625421166419982, "loss_raw": 2.1774001121520996, "perplexity": 6.440087433419259, "muon_lr": 0.005155028929232151, "adamw_lr": 0.00015465086787696449, "tokens_per_sec": 150038.7449780297, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 5217976320, "tokens_processed_billions": 5.21797632}
+{"timestamp": 1787561416.7347524, "elapsed_hours": 9.664679512514008, "session_id": 2, "event": "train_step", "step": 79630, "epoch": 1, "loss": 1.909319806098938, "loss_raw": 2.901428461074829, "perplexity": 6.748496950567725, "muon_lr": 0.00515426701177015, "adamw_lr": 0.00015462801035310444, "tokens_per_sec": 150039.97772766906, "vram_gb": 27.710740089416504, "grad_norm": 0.134765625, "tokens_processed": 5218631680, "tokens_processed_billions": 5.21863168}
+{"timestamp": 1787561420.8613143, "elapsed_hours": 9.665825779702928, "session_id": 2, "event": "train_step", "step": 79640, "epoch": 1, "loss": 1.906866431236267, "loss_raw": 2.4654381275177, "perplexity": 6.7319606509536865, "muon_lr": 0.005152743206778256, "adamw_lr": 0.00015458229620334763, "tokens_per_sec": 150041.0184133927, "vram_gb": 27.299458026885986, "grad_norm": 0.1376953125, "tokens_processed": 5219287040, "tokens_processed_billions": 5.21928704}
+{"timestamp": 1787561424.9171855, "elapsed_hours": 9.666952410538991, "session_id": 2, "event": "train_step", "step": 79650, "epoch": 1, "loss": 1.8256886649131774, "loss_raw": 0.9971103668212891, "perplexity": 6.207068137602663, "muon_lr": 0.0051519813192923035, "adamw_lr": 0.00015455943957876905, "tokens_per_sec": 150042.36413838653, "vram_gb": 27.710740089416504, "grad_norm": 34.25, "tokens_processed": 5219942400, "tokens_processed_billions": 5.2199424}
+{"timestamp": 1787561428.89492, "elapsed_hours": 9.668057336939706, "session_id": 2, "event": "train_step", "step": 79660, "epoch": 1, "loss": 1.8047421967983246, "loss_raw": 0.9193997383117676, "perplexity": 6.078404214442477, "muon_lr": 0.00515121944184231, "adamw_lr": 0.00015453658325526926, "tokens_per_sec": 150044.04636639924, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 5220597760, "tokens_processed_billions": 5.22059776}
+{"timestamp": 1787561433.0345166, "elapsed_hours": 9.669207224845886, "session_id": 2, "event": "train_step", "step": 79670, "epoch": 1, "loss": 1.7575950884819032, "loss_raw": 1.6527197360992432, "perplexity": 5.79847579170317, "muon_lr": 0.005150457574450245, "adamw_lr": 0.0001545137272335073, "tokens_per_sec": 150045.0307927096, "vram_gb": 27.710740089416504, "grad_norm": 0.1318359375, "tokens_processed": 5221253120, "tokens_processed_billions": 5.22125312}
+{"timestamp": 1787561437.1752732, "elapsed_hours": 9.670357435146968, "session_id": 2, "event": "train_step", "step": 79680, "epoch": 1, "loss": 1.7633308899402618, "loss_raw": 3.7530465126037598, "perplexity": 5.831830263474215, "muon_lr": 0.005148933869927783, "adamw_lr": 0.00015446801609783346, "tokens_per_sec": 150046.0092973506, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 5221908480, "tokens_processed_billions": 5.22190848}
+{"timestamp": 1787561441.1632595, "elapsed_hours": 9.671465209060244, "session_id": 2, "event": "train_step", "step": 79690, "epoch": 1, "loss": 1.7269282293319703, "loss_raw": 1.4913324117660522, "perplexity": 5.623353697946864, "muon_lr": 0.005148172032841325, "adamw_lr": 0.0001544451609852397, "tokens_per_sec": 150047.64642308315, "vram_gb": 27.710740089416504, "grad_norm": 0.1337890625, "tokens_processed": 5222563840, "tokens_processed_billions": 5.22256384}
+{"timestamp": 1787561445.344205, "elapsed_hours": 9.672626582715246, "session_id": 2, "event": "train_step", "step": 79700, "epoch": 1, "loss": 1.7092426538467407, "loss_raw": 2.2594187259674072, "perplexity": 5.524775725149244, "muon_lr": 0.005147410205900675, "adamw_lr": 0.0001544223061770202, "tokens_per_sec": 150048.45127344056, "vram_gb": 27.710740089416504, "grad_norm": 0.126953125, "tokens_processed": 5223219200, "tokens_processed_billions": 5.2232192}
+{"timestamp": 1787561445.3496077, "elapsed_hours": 9.67262808362643, "session_id": 2, "event": "lambda_stats", "step": 79700, "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": 1787561449.3795052, "elapsed_hours": 9.673747499386469, "session_id": 2, "event": "train_step", "step": 79710, "epoch": 1, "loss": 1.7546234762668609, "loss_raw": 2.024397373199463, "perplexity": 5.781270546531155, "muon_lr": 0.005146648389127801, "adamw_lr": 0.00015439945167383402, "tokens_per_sec": 150049.88353290025, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 5223874560, "tokens_processed_billions": 5.22387456}
+{"timestamp": 1787561453.5077662, "elapsed_hours": 9.674894238577949, "session_id": 2, "event": "train_step", "step": 79720, "epoch": 1, "loss": 1.7481466591358186, "loss_raw": 2.289246082305908, "perplexity": 5.7439473127692, "muon_lr": 0.0051451247861732646, "adamw_lr": 0.0001543537435851979, "tokens_per_sec": 150050.91433715314, "vram_gb": 27.299458026885986, "grad_norm": 0.1318359375, "tokens_processed": 5224529920, "tokens_processed_billions": 5.22452992}
+{"timestamp": 1787561457.562138, "elapsed_hours": 9.676020453174909, "session_id": 2, "event": "train_step", "step": 79730, "epoch": 1, "loss": 1.7518254971504212, "loss_raw": 1.0351265668869019, "perplexity": 5.76511728106343, "muon_lr": 0.005144363000035535, "adamw_lr": 0.000154330890001066, "tokens_per_sec": 150052.26449014244, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 5225185280, "tokens_processed_billions": 5.22518528}
+{"timestamp": 1787561461.6637833, "elapsed_hours": 9.677159799006251, "session_id": 2, "event": "train_step", "step": 79740, "epoch": 1, "loss": 1.785108687877655, "loss_raw": 1.5827789306640625, "perplexity": 5.960227716880502, "muon_lr": 0.005143601224153457, "adamw_lr": 0.0001543080367246037, "tokens_per_sec": 150053.41022750683, "vram_gb": 27.710740089416504, "grad_norm": 0.130859375, "tokens_processed": 5225840640, "tokens_processed_billions": 5.22584064}
+{"timestamp": 1787561465.6502388, "elapsed_hours": 9.6782671476735, "session_id": 2, "event": "train_step", "step": 79750, "epoch": 1, "loss": 1.8394606351852416, "loss_raw": 1.7630332708358765, "perplexity": 6.293143045546073, "muon_lr": 0.005142839458548996, "adamw_lr": 0.00015428518375646986, "tokens_per_sec": 150055.0518032347, "vram_gb": 27.710740089416504, "grad_norm": 0.138671875, "tokens_processed": 5226496000, "tokens_processed_billions": 5.226496}
+{"timestamp": 1787561469.9705167, "elapsed_hours": 9.679467224942313, "session_id": 2, "event": "train_step", "step": 79760, "epoch": 1, "loss": 1.8483607518672942, "loss_raw": 2.706045627593994, "perplexity": 6.3494027405029705, "muon_lr": 0.005141315958260803, "adamw_lr": 0.00015423947874782404, "tokens_per_sec": 150055.25519916633, "vram_gb": 27.299458026885986, "grad_norm": 0.140625, "tokens_processed": 5227151360, "tokens_processed_billions": 5.22715136}
+{"timestamp": 1787561473.9256077, "elapsed_hours": 9.680565861264865, "session_id": 2, "event": "train_step", "step": 79770, "epoch": 1, "loss": 1.9376074802875518, "loss_raw": 1.9694477319717407, "perplexity": 6.942121922361168, "muon_lr": 0.0051405542236210005, "adamw_lr": 0.00015421662670862997, "tokens_per_sec": 150057.03140733662, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 5227806720, "tokens_processed_billions": 5.22780672}
+{"timestamp": 1787561478.0310192, "elapsed_hours": 9.681706253555086, "session_id": 2, "event": "train_step", "step": 79780, "epoch": 1, "loss": 1.8565728557109833, "loss_raw": 1.653348684310913, "perplexity": 6.401759379995504, "muon_lr": 0.005139792499346686, "adamw_lr": 0.00015419377498040052, "tokens_per_sec": 150058.15978816603, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 5228462080, "tokens_processed_billions": 5.22846208}
+{"timestamp": 1787561482.1289823, "elapsed_hours": 9.682844576438269, "session_id": 2, "event": "train_step", "step": 79790, "epoch": 1, "loss": 1.9473203909397125, "loss_raw": 2.138239622116089, "perplexity": 7.009878657201661, "muon_lr": 0.0051390307854598215, "adamw_lr": 0.0001541709235637946, "tokens_per_sec": 150059.3197293547, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 5229117440, "tokens_processed_billions": 5.22911744}
+{"timestamp": 1787561486.2772229, "elapsed_hours": 9.683996865616905, "session_id": 2, "event": "train_step", "step": 79800, "epoch": 1, "loss": 1.9654273974895478, "loss_raw": 1.5451996326446533, "perplexity": 7.137962685642764, "muon_lr": 0.005137507388936317, "adamw_lr": 0.00015412522166808945, "tokens_per_sec": 150060.26324486366, "vram_gb": 27.299458026885986, "grad_norm": 0.14453125, "tokens_processed": 5229772800, "tokens_processed_billions": 5.2297728}
+{"timestamp": 1787561486.280919, "elapsed_hours": 9.683997892273798, "session_id": 2, "event": "lambda_stats", "step": 79800, "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": 1787561490.3968651, "elapsed_hours": 9.685141210556031, "session_id": 2, "event": "train_step", "step": 79810, "epoch": 1, "loss": 1.9657164311408997, "loss_raw": 1.5714662075042725, "perplexity": 7.14002609524405, "muon_lr": 0.005136745706343603, "adamw_lr": 0.00015410237119030805, "tokens_per_sec": 150061.32960368777, "vram_gb": 27.710740089416504, "grad_norm": 0.1650390625, "tokens_processed": 5230428160, "tokens_processed_billions": 5.23042816}
+{"timestamp": 1787561494.4645782, "elapsed_hours": 9.686271130707528, "session_id": 2, "event": "train_step", "step": 79820, "epoch": 1, "loss": 1.9254756200313567, "loss_raw": 2.2907657623291016, "perplexity": 6.85840988741227, "muon_lr": 0.005135984034226205, "adamw_lr": 0.00015407952102678611, "tokens_per_sec": 150062.61864977903, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 5231083520, "tokens_processed_billions": 5.23108352}
+{"timestamp": 1787561498.449287, "elapsed_hours": 9.687377994325425, "session_id": 2, "event": "train_step", "step": 79830, "epoch": 1, "loss": 1.8693678402900695, "loss_raw": 0.9499677419662476, "perplexity": 6.484196055889492, "muon_lr": 0.005135222372606084, "adamw_lr": 0.0001540566711781825, "tokens_per_sec": 150064.26566047606, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 5231738880, "tokens_processed_billions": 5.23173888}
+{"timestamp": 1787561502.6723592, "elapsed_hours": 9.688551070094109, "session_id": 2, "event": "train_step", "step": 79840, "epoch": 1, "loss": 1.8544039392471314, "loss_raw": 1.0570489168167114, "perplexity": 6.3878895453721505, "muon_lr": 0.005133699080945537, "adamw_lr": 0.00015401097242836607, "tokens_per_sec": 150064.88564419845, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 5232394240, "tokens_processed_billions": 5.23239424}
+{"timestamp": 1787561506.6487565, "elapsed_hours": 9.689655624826749, "session_id": 2, "event": "train_step", "step": 79850, "epoch": 1, "loss": 1.8623242282867432, "loss_raw": 2.3017940521240234, "perplexity": 6.438684366222246, "muon_lr": 0.005132937450949037, "adamw_lr": 0.0001539881235284711, "tokens_per_sec": 150066.56754109517, "vram_gb": 27.710740089416504, "grad_norm": 0.14453125, "tokens_processed": 5233049600, "tokens_processed_billions": 5.2330496}
+{"timestamp": 1787561510.6128404, "elapsed_hours": 9.69075675924619, "session_id": 2, "event": "train_step", "step": 79860, "epoch": 1, "loss": 1.7929701209068298, "loss_raw": 2.4024195671081543, "perplexity": 6.007268308880073, "muon_lr": 0.005132175831537673, "adamw_lr": 0.00015396527494613015, "tokens_per_sec": 150068.30159329527, "vram_gb": 27.710740089416504, "grad_norm": 0.1259765625, "tokens_processed": 5233704960, "tokens_processed_billions": 5.23370496}
+{"timestamp": 1787561514.5883179, "elapsed_hours": 9.691861058606042, "session_id": 2, "event": "train_step", "step": 79870, "epoch": 1, "loss": 1.8777499115467071, "loss_raw": 1.104772686958313, "perplexity": 6.538775473999729, "muon_lr": 0.005131414222733403, "adamw_lr": 0.00015394242668200205, "tokens_per_sec": 150069.98650821715, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 5234360320, "tokens_processed_billions": 5.23436032}
+{"timestamp": 1787561518.7127209, "elapsed_hours": 9.693006726238462, "session_id": 2, "event": "train_step", "step": 79880, "epoch": 1, "loss": 1.991782169342041, "loss_raw": 2.3067517280578613, "perplexity": 7.328582906740181, "muon_lr": 0.0051298910370340105, "adamw_lr": 0.00015389673111102027, "tokens_per_sec": 150071.03013438662, "vram_gb": 27.299458026885986, "grad_norm": 0.154296875, "tokens_processed": 5235015680, "tokens_processed_billions": 5.23501568}
+{"timestamp": 1787561522.7939932, "elapsed_hours": 9.694140412741238, "session_id": 2, "event": "train_step", "step": 79890, "epoch": 1, "loss": 1.939603604078293, "loss_raw": 1.6119120121002197, "perplexity": 6.955993096773424, "muon_lr": 0.005129129460182808, "adamw_lr": 0.0001538738838054842, "tokens_per_sec": 150072.25972464343, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 5235671040, "tokens_processed_billions": 5.23567104}
+{"timestamp": 1787561526.7694988, "elapsed_hours": 9.695244719982147, "session_id": 2, "event": "train_step", "step": 79900, "epoch": 1, "loss": 1.9603803265094757, "loss_raw": 2.5452141761779785, "perplexity": 7.102027640956681, "muon_lr": 0.005128367894026554, "adamw_lr": 0.0001538510368207966, "tokens_per_sec": 150073.94334809761, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 5236326400, "tokens_processed_billions": 5.2363264}
+{"timestamp": 1787561526.77316, "elapsed_hours": 9.69524573703607, "session_id": 2, "event": "lambda_stats", "step": 79900, "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": 1787561530.7485347, "elapsed_hours": 9.696350007653237, "session_id": 2, "event": "train_step", "step": 79910, "epoch": 1, "loss": 1.9760024559497833, "loss_raw": 2.3350141048431396, "perplexity": 7.2138475948112, "muon_lr": 0.005127606338587207, "adamw_lr": 0.00015382819015761617, "tokens_per_sec": 150075.61097774783, "vram_gb": 27.710740089416504, "grad_norm": 0.1572265625, "tokens_processed": 5236981760, "tokens_processed_billions": 5.23698176}
+{"timestamp": 1787561534.993263, "elapsed_hours": 9.697529098921352, "session_id": 2, "event": "train_step", "step": 79920, "epoch": 1, "loss": 1.9087454569339752, "loss_raw": 2.3117356300354004, "perplexity": 6.744622069853184, "muon_lr": 0.005126083259947084, "adamw_lr": 0.00015378249779841248, "tokens_per_sec": 150076.1363786799, "vram_gb": 27.299458026885986, "grad_norm": 0.1416015625, "tokens_processed": 5237637120, "tokens_processed_billions": 5.23763712}
+{"timestamp": 1787561538.9561396, "elapsed_hours": 9.698629897899098, "session_id": 2, "event": "train_step", "step": 79930, "epoch": 1, "loss": 1.915176283121109, "loss_raw": 2.94380521774292, "perplexity": 6.788135325213989, "muon_lr": 0.005125321736790231, "adamw_lr": 0.00015375965210370686, "tokens_per_sec": 150077.87331177673, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 5238292480, "tokens_processed_billions": 5.23829248}
+{"timestamp": 1787561543.0013318, "elapsed_hours": 9.699753562344446, "session_id": 2, "event": "train_step", "step": 79940, "epoch": 1, "loss": 1.878511331677437, "loss_raw": 1.5209252834320068, "perplexity": 6.543756125219374, "muon_lr": 0.005124560224438128, "adamw_lr": 0.0001537368067331438, "tokens_per_sec": 150079.25597075262, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 5238947840, "tokens_processed_billions": 5.23894784}
+{"timestamp": 1787561547.1000612, "elapsed_hours": 9.700892098347346, "session_id": 2, "event": "train_step", "step": 79950, "epoch": 1, "loss": 1.8430879098176955, "loss_raw": 2.5079877376556396, "perplexity": 6.316011453598263, "muon_lr": 0.005123798722912736, "adamw_lr": 0.00015371396168738203, "tokens_per_sec": 150080.40834780296, "vram_gb": 27.710740089416504, "grad_norm": 0.1357421875, "tokens_processed": 5239603200, "tokens_processed_billions": 5.2396032}
+{"timestamp": 1787561551.3295674, "elapsed_hours": 9.702066961063279, "session_id": 2, "event": "train_step", "step": 79960, "epoch": 1, "loss": 1.7696197372674942, "loss_raw": 1.9690314531326294, "perplexity": 5.868621319049415, "muon_lr": 0.0051222757524299264, "adamw_lr": 0.00015366827257289774, "tokens_per_sec": 150080.99791632086, "vram_gb": 27.299458026885986, "grad_norm": 0.12890625, "tokens_processed": 5240258560, "tokens_processed_billions": 5.24025856}
+{"timestamp": 1787561555.4980233, "elapsed_hours": 9.703224865661728, "session_id": 2, "event": "train_step", "step": 79970, "epoch": 1, "loss": 1.7457429963350295, "loss_raw": 1.157478928565979, "perplexity": 5.730157380097285, "muon_lr": 0.005121514283516427, "adamw_lr": 0.00015364542850549278, "tokens_per_sec": 150081.8502761758, "vram_gb": 27.710740089416504, "grad_norm": 0.1396484375, "tokens_processed": 5240913920, "tokens_processed_billions": 5.24091392}
+{"timestamp": 1787561559.5009513, "elapsed_hours": 9.704336789978875, "session_id": 2, "event": "train_step", "step": 79980, "epoch": 1, "loss": 1.6742645126581193, "loss_raw": 1.4460878372192383, "perplexity": 5.334869976454467, "muon_lr": 0.005120752825517477, "adamw_lr": 0.0001536225847655243, "tokens_per_sec": 150083.41337630918, "vram_gb": 27.710740089416504, "grad_norm": 0.134765625, "tokens_processed": 5241569280, "tokens_processed_billions": 5.24156928}
+{"timestamp": 1787561563.5743518, "elapsed_hours": 9.705468290249506, "session_id": 2, "event": "train_step", "step": 79990, "epoch": 1, "loss": 1.7120755577087403, "loss_raw": 2.1454007625579834, "perplexity": 5.540449073700522, "muon_lr": 0.005119991378455033, "adamw_lr": 0.00015359974135365094, "tokens_per_sec": 150084.67313303443, "vram_gb": 27.710740089416504, "grad_norm": 0.1396484375, "tokens_processed": 5242224640, "tokens_processed_billions": 5.24222464}
+{"timestamp": 1787561567.8118262, "elapsed_hours": 9.70664536635081, "session_id": 2, "event": "train_step", "step": 80000, "epoch": 1, "loss": 1.6485423064231872, "loss_raw": 1.4004781246185303, "perplexity": 5.199395175523209, "muon_lr": 0.005118468517227499, "adamw_lr": 0.00015355405551682492, "tokens_per_sec": 150085.22812219957, "vram_gb": 27.299458026885986, "grad_norm": 0.1279296875, "tokens_processed": 5242880000, "tokens_processed_billions": 5.24288}
+{"timestamp": 1787561567.8151762, "elapsed_hours": 9.706646296911769, "session_id": 2, "event": "lambda_stats", "step": 80000, "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": 1787561581.3407931, "elapsed_hours": 9.710403412977854, "session_id": 2, "event": "validation", "step": 80000, "epoch": 1, "val_loss": 1.837772669196129, "val_perplexity": 6.282529394381723, "is_best": true, "best_loss": 1.837772669196129, "best_perplexity": 6.282529394381723, "train_loss": 1.4004781246185303, "train_loss_avg": 1.6485423064231872}