Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -5844,3 +5844,32 @@
 {"timestamp": 1787549260.6211724, "elapsed_hours": 6.287981295784315, "session_id": 2, "event": "train_step", "step": 51730, "epoch": 1, "loss": 1.8332701683044434, "loss_raw": 0.7629439234733582, "perplexity": 6.254305886047308, "muon_lr": 0.007717356157636753, "adamw_lr": 0.00023152068472910253, "tokens_per_sec": 149838.813556266, "vram_gb": 27.710740089416504, "grad_norm": 0.166015625, "tokens_processed": 3390177280, "tokens_processed_billions": 3.39017728}
 {"timestamp": 1787549264.4260545, "elapsed_hours": 6.28903820739852, "session_id": 2, "event": "train_step", "step": 51740, "epoch": 1, "loss": 1.790153248310089, "loss_raw": 0.9172790050506592, "perplexity": 5.990370410186817, "muon_lr": 0.007716691164309394, "adamw_lr": 0.00023150073492928173, "tokens_per_sec": 149842.58030002876, "vram_gb": 27.710740089416504, "grad_norm": 0.138671875, "tokens_processed": 3390832640, "tokens_processed_billions": 3.39083264}
 {"timestamp": 1787549268.2355325, "elapsed_hours": 6.290096395810445, "session_id": 2, "event": "train_step", "step": 51750, "epoch": 1, "loss": 1.831990225315094, "loss_raw": 2.1079838275909424, "perplexity": 6.24630585196116, "muon_lr": 0.007716026107058397, "adamw_lr": 0.00023148078321175186, "tokens_per_sec": 149846.31548134913, "vram_gb": 27.710740089416504, "grad_norm": 0.16015625, "tokens_processed": 3391488000, "tokens_processed_billions": 3.391488}
+{"timestamp": 1787549272.9138942, "elapsed_hours": 6.291395940714413, "session_id": 2, "event": "train_step", "step": 51760, "epoch": 1, "loss": 1.886046986579895, "loss_raw": 1.2810419797897339, "perplexity": 6.59325387790443, "muon_lr": 0.007714695800862206, "adamw_lr": 0.0002314408740258661, "tokens_per_sec": 149844.2976711871, "vram_gb": 27.299458026885986, "grad_norm": 0.1982421875, "tokens_processed": 3392143360, "tokens_processed_billions": 3.39214336}
+{"timestamp": 1787549276.7195907, "elapsed_hours": 6.292453078561359, "session_id": 2, "event": "train_step", "step": 51770, "epoch": 1, "loss": 1.835250358581543, "loss_raw": 2.2393882274627686, "perplexity": 6.266702871897414, "muon_lr": 0.0077140305519553766, "adamw_lr": 0.00023142091655866122, "tokens_per_sec": 149848.0561606782, "vram_gb": 27.710740089416504, "grad_norm": 0.13671875, "tokens_processed": 3392798720, "tokens_processed_billions": 3.39279872}
+{"timestamp": 1787549280.520795, "elapsed_hours": 6.293508968684408, "session_id": 2, "event": "train_step", "step": 51780, "epoch": 1, "loss": 1.7943204391002654, "loss_raw": 2.125702381134033, "perplexity": 6.015385511740372, "muon_lr": 0.007713365239201634, "adamw_lr": 0.00023140095717604897, "tokens_per_sec": 149851.84336146517, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 3393454080, "tokens_processed_billions": 3.39345408}
+{"timestamp": 1787549284.3250847, "elapsed_hours": 6.294565715789795, "session_id": 2, "event": "train_step", "step": 51790, "epoch": 1, "loss": 1.8715125453472137, "loss_raw": 1.8658679723739624, "perplexity": 6.4981176675011465, "muon_lr": 0.007712699862620167, "adamw_lr": 0.00023138099587860495, "tokens_per_sec": 149855.60859311037, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 3394109440, "tokens_processed_billions": 3.39410944}
+{"timestamp": 1787549288.219674, "elapsed_hours": 6.295647546251615, "session_id": 2, "event": "train_step", "step": 51800, "epoch": 1, "loss": 1.904249790906906, "loss_raw": 2.62858247756958, "perplexity": 6.714368557347004, "muon_lr": 0.007711368918050804, "adamw_lr": 0.00023134106754152406, "tokens_per_sec": 149858.77496457178, "vram_gb": 27.299458026885986, "grad_norm": 0.1474609375, "tokens_processed": 3394764800, "tokens_processed_billions": 3.3947648}
+{"timestamp": 1787549288.2212722, "elapsed_hours": 6.295647990107536, "session_id": 2, "event": "lambda_stats", "step": 51800, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1901041666666667, "lambda_std": 0.20448927530625605, "lambda_all": [2.015625, 1.25, 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": 1787549292.022591, "elapsed_hours": 6.296703912085957, "session_id": 2, "event": "train_step", "step": 51810, "epoch": 1, "loss": 1.9642088329792022, "loss_raw": 0.9936015605926514, "perplexity": 7.129269915064389, "muon_lr": 0.007710703350101292, "adamw_lr": 0.0002313211005030387, "tokens_per_sec": 149862.54712187554, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 3395420160, "tokens_processed_billions": 3.39542016}
+{"timestamp": 1787549295.8378553, "elapsed_hours": 6.297763707704014, "session_id": 2, "event": "train_step", "step": 51820, "epoch": 1, "loss": 2.061590163707733, "loss_raw": 1.7766690254211426, "perplexity": 7.8584561117633935, "muon_lr": 0.007710037718400815, "adamw_lr": 0.0002313011315520244, "tokens_per_sec": 149866.23616097527, "vram_gb": 27.710740089416504, "grad_norm": 0.15234375, "tokens_processed": 3396075520, "tokens_processed_billions": 3.39607552}
+{"timestamp": 1787549299.654537, "elapsed_hours": 6.298823896977637, "session_id": 2, "event": "train_step", "step": 51830, "epoch": 1, "loss": 2.107494617700577, "loss_raw": 1.5435234308242798, "perplexity": 8.227602152165241, "muon_lr": 0.00770937202296857, "adamw_lr": 0.00023128116068905704, "tokens_per_sec": 149869.91451296484, "vram_gb": 27.710740089416504, "grad_norm": 0.1650390625, "tokens_processed": 3396730880, "tokens_processed_billions": 3.39673088}
+{"timestamp": 1787549303.5464947, "elapsed_hours": 6.299904996355375, "session_id": 2, "event": "train_step", "step": 51840, "epoch": 1, "loss": 2.0366007351875304, "loss_raw": 1.5159035921096802, "perplexity": 7.6645111638391175, "muon_lr": 0.007708040440985561, "adamw_lr": 0.00023124121322956677, "tokens_per_sec": 149873.09363104793, "vram_gb": 27.299458026885986, "grad_norm": 0.1328125, "tokens_processed": 3397386240, "tokens_processed_billions": 3.39738624}
+{"timestamp": 1787549307.354784, "elapsed_hours": 6.300962854557567, "session_id": 2, "event": "train_step", "step": 51850, "epoch": 1, "loss": 2.049512403011322, "loss_raw": 0.9399539232254028, "perplexity": 7.764114424381365, "muon_lr": 0.007707374554473197, "adamw_lr": 0.00023122123663419584, "tokens_per_sec": 149876.82533206008, "vram_gb": 27.710740089416504, "grad_norm": 0.15625, "tokens_processed": 3398041600, "tokens_processed_billions": 3.3980416}
+{"timestamp": 1787549311.1699517, "elapsed_hours": 6.302022623353534, "session_id": 2, "event": "train_step", "step": 51860, "epoch": 1, "loss": 2.013212615251541, "loss_raw": 3.29599666595459, "perplexity": 7.487332669796945, "muon_lr": 0.007706708604305863, "adamw_lr": 0.00023120125812917583, "tokens_per_sec": 149880.51041059752, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 3398696960, "tokens_processed_billions": 3.39869696}
+{"timestamp": 1787549314.979834, "elapsed_hours": 6.303080923954646, "session_id": 2, "event": "train_step", "step": 51870, "epoch": 1, "loss": 1.8857852971553803, "loss_raw": 2.7930727005004883, "perplexity": 6.591528718829229, "muon_lr": 0.007706042590502762, "adamw_lr": 0.0002311812777150828, "tokens_per_sec": 149884.2284453694, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 3399352320, "tokens_processed_billions": 3.39935232}
+{"timestamp": 1787549318.8772209, "elapsed_hours": 6.304163531396124, "session_id": 2, "event": "train_step", "step": 51880, "epoch": 1, "loss": 1.9119617867469787, "loss_raw": 1.0888118743896484, "perplexity": 6.76634992213261, "muon_lr": 0.007704710372066091, "adamw_lr": 0.00023114131116198266, "tokens_per_sec": 149887.3671587525, "vram_gb": 27.299458026885986, "grad_norm": 0.1572265625, "tokens_processed": 3400007680, "tokens_processed_billions": 3.40000768}
+{"timestamp": 1787549322.6898558, "elapsed_hours": 6.3052225966586, "session_id": 2, "event": "train_step", "step": 51890, "epoch": 1, "loss": 1.9927998435497285, "loss_raw": 2.79325795173645, "perplexity": 7.336044812793935, "muon_lr": 0.007704044167470937, "adamw_lr": 0.00023112132502412804, "tokens_per_sec": 149891.06512760464, "vram_gb": 27.710740089416504, "grad_norm": 0.16015625, "tokens_processed": 3400663040, "tokens_processed_billions": 3.40066304}
+{"timestamp": 1787549326.501738, "elapsed_hours": 6.306281452841229, "session_id": 2, "event": "train_step", "step": 51900, "epoch": 1, "loss": 1.9384832239151002, "loss_raw": 1.528182864189148, "perplexity": 6.948204104223628, "muon_lr": 0.007703377899316852, "adamw_lr": 0.0002311013369795055, "tokens_per_sec": 149894.76672352344, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 3401318400, "tokens_processed_billions": 3.4013184}
+{"timestamp": 1787549326.5035908, "elapsed_hours": 6.306281967494223, "session_id": 2, "event": "lambda_stats", "step": 51900, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1901041666666667, "lambda_std": 0.20448927530625605, "lambda_all": [2.015625, 1.25, 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": 1787549330.3225303, "elapsed_hours": 6.307342784073618, "session_id": 2, "event": "train_step", "step": 51910, "epoch": 1, "loss": 1.9321799886226654, "loss_raw": 2.8344428539276123, "perplexity": 6.90454567810949, "muon_lr": 0.00770271156762305, "adamw_lr": 0.00023108134702869143, "tokens_per_sec": 149898.40823186794, "vram_gb": 27.710740089416504, "grad_norm": 0.1328125, "tokens_processed": 3401973760, "tokens_processed_billions": 3.40197376}
+{"timestamp": 1787549334.2186043, "elapsed_hours": 6.308425026867125, "session_id": 2, "event": "train_step", "step": 51920, "epoch": 1, "loss": 2.0136613750457766, "loss_raw": 1.886913537979126, "perplexity": 7.490693437697572, "muon_lr": 0.00770137871369316, "adamw_lr": 0.00023104136141079474, "tokens_per_sec": 149901.5510282642, "vram_gb": 27.299458026885986, "grad_norm": 0.1435546875, "tokens_processed": 3402629120, "tokens_processed_billions": 3.40262912}
+{"timestamp": 1787549338.0275424, "elapsed_hours": 6.309483065207799, "session_id": 2, "event": "train_step", "step": 51930, "epoch": 1, "loss": 1.832916383743286, "loss_raw": 1.1152515411376953, "perplexity": 6.252093600543371, "muon_lr": 0.007700712191495507, "adamw_lr": 0.00023102136574486515, "tokens_per_sec": 149905.26861642255, "vram_gb": 27.710740089416504, "grad_norm": 0.130859375, "tokens_processed": 3403284480, "tokens_processed_billions": 3.40328448}
+{"timestamp": 1787549341.842718, "elapsed_hours": 6.31054283618927, "session_id": 2, "event": "train_step", "step": 51940, "epoch": 1, "loss": 1.848555953502655, "loss_raw": 2.3362271785736084, "perplexity": 6.350642275277153, "muon_lr": 0.007700045605835008, "adamw_lr": 0.0002310013681750502, "tokens_per_sec": 149908.94340452133, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 3403939840, "tokens_processed_billions": 3.40393984}
+{"timestamp": 1787549345.651933, "elapsed_hours": 6.311600951419936, "session_id": 2, "event": "train_step", "step": 51950, "epoch": 1, "loss": 1.8962537634372711, "loss_raw": 1.3545246124267578, "perplexity": 6.660894357419237, "muon_lr": 0.007699378956730888, "adamw_lr": 0.00023098136870192658, "tokens_per_sec": 149912.65651240537, "vram_gb": 27.710740089416504, "grad_norm": 0.1533203125, "tokens_processed": 3404595200, "tokens_processed_billions": 3.4045952}
+{"timestamp": 1787549349.5435221, "elapsed_hours": 6.31268194841014, "session_id": 2, "event": "train_step", "step": 51960, "epoch": 1, "loss": 1.8592166936397552, "loss_raw": 1.610605001449585, "perplexity": 6.418706987747486, "muon_lr": 0.007698045468268679, "adamw_lr": 0.0002309413640480603, "tokens_per_sec": 149915.82432853148, "vram_gb": 27.299458026885986, "grad_norm": 0.171875, "tokens_processed": 3405250560, "tokens_processed_billions": 3.40525056}
+{"timestamp": 1787549353.3537164, "elapsed_hours": 6.313740335901579, "session_id": 2, "event": "train_step", "step": 51970, "epoch": 1, "loss": 1.912027313709259, "loss_raw": 1.4729132652282715, "perplexity": 6.7667933150156685, "muon_lr": 0.007697378628949045, "adamw_lr": 0.0002309213588684713, "tokens_per_sec": 149919.5287196496, "vram_gb": 27.710740089416504, "grad_norm": 0.16015625, "tokens_processed": 3405905920, "tokens_processed_billions": 3.40590592}
+{"timestamp": 1787549357.1666334, "elapsed_hours": 6.314799479378594, "session_id": 2, "event": "train_step", "step": 51980, "epoch": 1, "loss": 2.0053935432434082, "loss_raw": 1.1704061031341553, "perplexity": 7.429016960974169, "muon_lr": 0.007696711726262697, "adamw_lr": 0.00023090135178788085, "tokens_per_sec": 149923.21365799106, "vram_gb": 27.710740089416504, "grad_norm": 0.138671875, "tokens_processed": 3406561280, "tokens_processed_billions": 3.40656128}
+{"timestamp": 1787549360.9741702, "elapsed_hours": 6.315857128434711, "session_id": 2, "event": "train_step", "step": 51990, "epoch": 1, "loss": 1.9563413166999817, "loss_raw": 2.069262981414795, "perplexity": 7.07340033355313, "muon_lr": 0.007696044760228868, "adamw_lr": 0.00023088134280686598, "tokens_per_sec": 149926.9329654332, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 3407216640, "tokens_processed_billions": 3.40721664}
+{"timestamp": 1787549364.879676, "elapsed_hours": 6.3169419911834925, "session_id": 2, "event": "train_step", "step": 52000, "epoch": 1, "loss": 1.9297963690757751, "loss_raw": 2.048290491104126, "perplexity": 6.8881074670723255, "muon_lr": 0.0076947106381957, "adamw_lr": 0.0002308413191458709, "tokens_per_sec": 149930.0044482359, "vram_gb": 27.299458026885986, "grad_norm": 0.142578125, "tokens_processed": 3407872000, "tokens_processed_billions": 3.407872}
+{"timestamp": 1787549364.8815064, "elapsed_hours": 6.3169424997435675, "session_id": 2, "event": "lambda_stats", "step": 52000, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1901041666666667, "lambda_std": 0.20448927530625605, "lambda_all": [2.015625, 1.25, 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": 1787549378.0772867, "elapsed_hours": 6.320607994132572, "session_id": 2, "event": "validation", "step": 52000, "epoch": 1, "val_loss": 1.9220162492990494, "val_perplexity": 6.834725095835118, "is_best": true, "best_loss": 1.9220162492990494, "best_perplexity": 6.834725095835118, "train_loss": 2.048290491104126, "train_loss_avg": 1.9297963690757751}