Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -2964,3 +2964,32 @@
 {"timestamp": 1787648487.8291342, "elapsed_hours": 5.8974783977535035, "session_id": 2, "event": "train_step", "step": 120230, "epoch": 1, "loss": 1.6871248805522918, "loss_raw": 1.7320891618728638, "perplexity": 5.403921428866446, "muon_lr": 0.0018693870519958895, "adamw_lr": 5.6081611559876674e-05, "tokens_per_sec": 745023.5663463699, "vram_gb": 2.785593032836914, "grad_norm": 0.10400390625, "tokens_processed": 15758786560, "tokens_processed_billions": 15.75878656}
 {"timestamp": 1787648495.3764427, "elapsed_hours": 5.899574872321553, "session_id": 2, "event": "train_step", "step": 120240, "epoch": 1, "loss": 1.6617318093776703, "loss_raw": 1.5629445314407349, "perplexity": 5.268426859609299, "muon_lr": 0.0018684842850754044, "adamw_lr": 5.605452855226212e-05, "tokens_per_sec": 744819.7711375686, "vram_gb": 2.363645076751709, "grad_norm": 0.0947265625, "tokens_processed": 15760097280, "tokens_processed_billions": 15.76009728}
 {"timestamp": 1787648502.8494642, "elapsed_hours": 5.901650711562898, "session_id": 2, "event": "train_step", "step": 120250, "epoch": 1, "loss": 1.6249483287334443, "loss_raw": 1.650846004486084, "perplexity": 5.078156635615699, "muon_lr": 0.001868033058693929, "adamw_lr": 5.6040991760817864e-05, "tokens_per_sec": 744618.7375203395, "vram_gb": 2.785593032836914, "grad_norm": 0.08984375, "tokens_processed": 15761408000, "tokens_processed_billions": 15.761408}
+{"timestamp": 1787648510.954882, "elapsed_hours": 5.90390221648746, "session_id": 2, "event": "train_step", "step": 120260, "epoch": 1, "loss": 1.6556851720809938, "loss_raw": 1.218788504600525, "perplexity": 5.236666711081841, "muon_lr": 0.0018675819370489821, "adamw_lr": 5.602745811146945e-05, "tokens_per_sec": 744395.6121072485, "vram_gb": 2.785593032836914, "grad_norm": 0.09326171875, "tokens_processed": 15762718720, "tokens_processed_billions": 15.76271872}
+{"timestamp": 1787648518.426979, "elapsed_hours": 5.9059777990314695, "session_id": 2, "event": "train_step", "step": 120270, "epoch": 1, "loss": 1.6203686094284058, "loss_raw": 1.222309947013855, "perplexity": 5.054953276628367, "muon_lr": 0.0018671309201535712, "adamw_lr": 5.6013927604607133e-05, "tokens_per_sec": 744194.9068906596, "vram_gb": 2.785593032836914, "grad_norm": 0.09130859375, "tokens_processed": 15764029440, "tokens_processed_billions": 15.76402944}
+{"timestamp": 1787648525.9786766, "elapsed_hours": 5.908075492713187, "session_id": 2, "event": "train_step", "step": 120280, "epoch": 1, "loss": 1.6474874138832092, "loss_raw": 2.1258668899536133, "perplexity": 5.193913264262155, "muon_lr": 0.001866229200663385, "adamw_lr": 5.5986876019901546e-05, "tokens_per_sec": 743991.5465045722, "vram_gb": 2.363645076751709, "grad_norm": 0.095703125, "tokens_processed": 15765340160, "tokens_processed_billions": 15.76534016}
+{"timestamp": 1787648533.4566274, "elapsed_hours": 5.910152701470587, "session_id": 2, "event": "train_step", "step": 120290, "epoch": 1, "loss": 1.6616224586963653, "loss_raw": 2.184180974960327, "perplexity": 5.267850785040443, "muon_lr": 0.001865778498094613, "adamw_lr": 5.597335494283839e-05, "tokens_per_sec": 743790.922364375, "vram_gb": 2.785593032836914, "grad_norm": 0.09228515625, "tokens_processed": 15766650880, "tokens_processed_billions": 15.76665088}
+{"timestamp": 1787648540.9134862, "elapsed_hours": 5.912224050958951, "session_id": 2, "event": "train_step", "step": 120300, "epoch": 1, "loss": 1.7597108376026154, "loss_raw": 1.4116475582122803, "perplexity": 5.810756899050717, "muon_lr": 0.0018653279003273849, "adamw_lr": 5.595983700982154e-05, "tokens_per_sec": 743591.1766340149, "vram_gb": 2.785593032836914, "grad_norm": 0.103515625, "tokens_processed": 15767961600, "tokens_processed_billions": 15.7679616}
+{"timestamp": 1787648540.915264, "elapsed_hours": 5.912224544882775, "session_id": 2, "event": "lambda_stats", "step": 120300, "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": 1787648548.3991315, "elapsed_hours": 5.914303397006459, "session_id": 2, "event": "train_step", "step": 120310, "epoch": 1, "loss": 1.7742167365550996, "loss_raw": 1.5795106887817383, "perplexity": 5.895661471233877, "muon_lr": 0.0018648774073746937, "adamw_lr": 5.59463222212408e-05, "tokens_per_sec": 743390.562549097, "vram_gb": 2.785593032836914, "grad_norm": 0.0966796875, "tokens_processed": 15769272320, "tokens_processed_billions": 15.76927232}
+{"timestamp": 1787648555.9834623, "elapsed_hours": 5.91641015542878, "session_id": 2, "event": "train_step", "step": 120320, "epoch": 1, "loss": 1.7842076373100282, "loss_raw": 1.9142624139785767, "perplexity": 5.954859669117454, "muon_lr": 0.001863976735964888, "adamw_lr": 5.591930207894663e-05, "tokens_per_sec": 743186.6355949413, "vram_gb": 2.363645076751709, "grad_norm": 0.0908203125, "tokens_processed": 15770583040, "tokens_processed_billions": 15.77058304}
+{"timestamp": 1787648563.460572, "elapsed_hours": 5.918487130403519, "session_id": 2, "event": "train_step", "step": 120330, "epoch": 1, "loss": 1.746979157924652, "loss_raw": 1.4356688261032104, "perplexity": 5.737245160471414, "muon_lr": 0.0018635265575337467, "adamw_lr": 5.5905796726012395e-05, "tokens_per_sec": 742986.6071085446, "vram_gb": 2.785593032836914, "grad_norm": 0.09423828125, "tokens_processed": 15771893760, "tokens_processed_billions": 15.77189376}
+{"timestamp": 1787648570.9260368, "elapsed_hours": 5.920560870634185, "session_id": 2, "event": "train_step", "step": 120340, "epoch": 1, "loss": 1.676591967344284, "loss_raw": 1.5350152254104614, "perplexity": 5.347301105414256, "muon_lr": 0.0018630764839690896, "adamw_lr": 5.5892294519072677e-05, "tokens_per_sec": 742787.124751535, "vram_gb": 2.785593032836914, "grad_norm": 0.08544921875, "tokens_processed": 15773204480, "tokens_processed_billions": 15.77320448}
+{"timestamp": 1787648578.4076114, "elapsed_hours": 5.922639085782899, "session_id": 2, "event": "train_step", "step": 120350, "epoch": 1, "loss": 1.660816229581833, "loss_raw": 3.281604528427124, "perplexity": 5.263605401972207, "muon_lr": 0.0018626265152838937, "adamw_lr": 5.58787954585168e-05, "tokens_per_sec": 742587.2223948677, "vram_gb": 2.785593032836914, "grad_norm": 0.09033203125, "tokens_processed": 15774515200, "tokens_processed_billions": 15.7745152}
+{"timestamp": 1787648585.9612043, "elapsed_hours": 5.924737305972311, "session_id": 2, "event": "train_step", "step": 120360, "epoch": 1, "loss": 1.7017491090297698, "loss_raw": 2.0888919830322266, "perplexity": 5.483530300871864, "muon_lr": 0.0018617268926037957, "adamw_lr": 5.585180677811386e-05, "tokens_per_sec": 742384.9386727387, "vram_gb": 2.363645076751709, "grad_norm": 0.0986328125, "tokens_processed": 15775825920, "tokens_processed_billions": 15.77582592}
+{"timestamp": 1787648593.4191418, "elapsed_hours": 5.926808955338267, "session_id": 2, "event": "train_step", "step": 120370, "epoch": 1, "loss": 1.6512861156463623, "loss_raw": 1.1335196495056152, "perplexity": 5.2136809136682825, "muon_lr": 0.001861277238634833, "adamw_lr": 5.583831715904499e-05, "tokens_per_sec": 742186.1432949231, "vram_gb": 2.785593032836914, "grad_norm": 0.08935546875, "tokens_processed": 15777136640, "tokens_processed_billions": 15.77713664}
+{"timestamp": 1787648600.8932846, "elapsed_hours": 5.9288851061132215, "session_id": 2, "event": "train_step", "step": 120380, "epoch": 1, "loss": 1.7095794200897216, "loss_raw": 2.5896835327148438, "perplexity": 5.526636596435243, "muon_lr": 0.0018608276895972205, "adamw_lr": 5.5824830687916605e-05, "tokens_per_sec": 741986.9231985596, "vram_gb": 2.785593032836914, "grad_norm": 0.09619140625, "tokens_processed": 15778447360, "tokens_processed_billions": 15.77844736}
+{"timestamp": 1787648608.3420193, "elapsed_hours": 5.930954199830691, "session_id": 2, "event": "train_step", "step": 120390, "epoch": 1, "loss": 1.7885756468772889, "loss_raw": 1.6446497440338135, "perplexity": 5.980927443821828, "muon_lr": 0.0018603782455039208, "adamw_lr": 5.581134736511761e-05, "tokens_per_sec": 741788.733825778, "vram_gb": 2.785593032836914, "grad_norm": 0.1025390625, "tokens_processed": 15779758080, "tokens_processed_billions": 15.77975808}
+{"timestamp": 1787648615.9047306, "elapsed_hours": 5.933054952157868, "session_id": 2, "event": "train_step", "step": 120400, "epoch": 1, "loss": 1.7328202629089355, "loss_raw": 1.2414144277572632, "perplexity": 5.656584489082109, "muon_lr": 0.001859479672202097, "adamw_lr": 5.57843901660629e-05, "tokens_per_sec": 741586.694425691, "vram_gb": 2.363645076751709, "grad_norm": 0.09423828125, "tokens_processed": 15781068800, "tokens_processed_billions": 15.7810688}
+{"timestamp": 1787648615.9064767, "elapsed_hours": 5.933055437339677, "session_id": 2, "event": "lambda_stats", "step": 120400, "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": 1787648623.3446763, "elapsed_hours": 5.935121603873041, "session_id": 2, "event": "train_step", "step": 120410, "epoch": 1, "loss": 1.7197242093086242, "loss_raw": 1.7427111864089966, "perplexity": 5.582988515671389, "muon_lr": 0.0018590305430194872, "adamw_lr": 5.577091629058461e-05, "tokens_per_sec": 741389.0846737769, "vram_gb": 2.785593032836914, "grad_norm": 0.09375, "tokens_processed": 15782379520, "tokens_processed_billions": 15.78237952}
+{"timestamp": 1787648630.8095813, "elapsed_hours": 5.9371951885355845, "session_id": 2, "event": "train_step", "step": 120420, "epoch": 1, "loss": 1.7623291897773743, "loss_raw": 1.0753114223480225, "perplexity": 5.82599144301129, "muon_lr": 0.0018585815188330152, "adamw_lr": 5.575744556499045e-05, "tokens_per_sec": 741190.745931097, "vram_gb": 2.785593032836914, "grad_norm": 0.09619140625, "tokens_processed": 15783690240, "tokens_processed_billions": 15.78369024}
+{"timestamp": 1787648638.2526186, "elapsed_hours": 5.939262698955006, "session_id": 2, "event": "train_step", "step": 120430, "epoch": 1, "loss": 1.7434698557853698, "loss_raw": 1.3514940738677979, "perplexity": 5.717146720132381, "muon_lr": 0.0018581325996556283, "adamw_lr": 5.5743977989668834e-05, "tokens_per_sec": 740993.3032563548, "vram_gb": 2.785593032836914, "grad_norm": 0.10009765625, "tokens_processed": 15785000960, "tokens_processed_billions": 15.78500096}
+{"timestamp": 1787648645.805353, "elapsed_hours": 5.941360680659612, "session_id": 2, "event": "train_step", "step": 120440, "epoch": 1, "loss": 1.7337450242042542, "loss_raw": 1.9780715703964233, "perplexity": 5.661817898935789, "muon_lr": 0.0018572350763798933, "adamw_lr": 5.5717052291396794e-05, "tokens_per_sec": 740792.1855373282, "vram_gb": 2.363645076751709, "grad_norm": 0.0947265625, "tokens_processed": 15786311680, "tokens_processed_billions": 15.78631168}
+{"timestamp": 1787648653.2438142, "elapsed_hours": 5.943426919910642, "session_id": 2, "event": "train_step", "step": 120450, "epoch": 1, "loss": 1.7327438342571257, "loss_raw": 1.6040959358215332, "perplexity": 5.656152180476352, "muon_lr": 0.0018567864723074284, "adamw_lr": 5.570359416922285e-05, "tokens_per_sec": 740595.1828695002, "vram_gb": 2.785593032836914, "grad_norm": 0.09619140625, "tokens_processed": 15787622400, "tokens_processed_billions": 15.7876224}
+{"timestamp": 1787648660.7218132, "elapsed_hours": 5.945504141847293, "session_id": 2, "event": "train_step", "step": 120460, "epoch": 1, "loss": 1.6889807975292206, "loss_raw": 1.1747699975967407, "perplexity": 5.413959970857101, "muon_lr": 0.0018563379732958127, "adamw_lr": 5.569013919887437e-05, "tokens_per_sec": 740396.9436376955, "vram_gb": 2.785593032836914, "grad_norm": 0.0908203125, "tokens_processed": 15788933120, "tokens_processed_billions": 15.78893312}
+{"timestamp": 1787648668.1615372, "elapsed_hours": 5.947570731904771, "session_id": 2, "event": "train_step", "step": 120470, "epoch": 1, "loss": 1.6744593727588652, "loss_raw": 1.5857131481170654, "perplexity": 5.335909631045856, "muon_lr": 0.001855889579357983, "adamw_lr": 5.5676687380739486e-05, "tokens_per_sec": 740200.17001287, "vram_gb": 2.785593032836914, "grad_norm": 0.0927734375, "tokens_processed": 15790243840, "tokens_processed_billions": 15.79024384}
+{"timestamp": 1787648675.7181787, "elapsed_hours": 5.949669798877504, "session_id": 2, "event": "train_step", "step": 120480, "epoch": 1, "loss": 1.6375144374370576, "loss_raw": 1.8894329071044922, "perplexity": 5.142371927077376, "muon_lr": 0.0018549931067553937, "adamw_lr": 5.5649793202661806e-05, "tokens_per_sec": 739999.4799863434, "vram_gb": 2.363645076751709, "grad_norm": 0.09375, "tokens_processed": 15791554560, "tokens_processed_billions": 15.79155456}
+{"timestamp": 1787648683.1762433, "elapsed_hours": 5.951741483476427, "session_id": 2, "event": "train_step", "step": 120490, "epoch": 1, "loss": 1.6351829183101654, "loss_raw": 2.5255422592163086, "perplexity": 5.130396354634819, "muon_lr": 0.0018545450281164874, "adamw_lr": 5.563635084349462e-05, "tokens_per_sec": 739802.3499009237, "vram_gb": 2.785593032836914, "grad_norm": 0.091796875, "tokens_processed": 15792865280, "tokens_processed_billions": 15.79286528}
+{"timestamp": 1787648690.6277814, "elapsed_hours": 5.953811355166965, "session_id": 2, "event": "train_step", "step": 120500, "epoch": 1, "loss": 1.6548359680175782, "loss_raw": 1.216089129447937, "perplexity": 5.232221600102322, "muon_lr": 0.0018540970546030687, "adamw_lr": 5.5622911638092046e-05, "tokens_per_sec": 739605.5815642223, "vram_gb": 2.785593032836914, "grad_norm": 0.10009765625, "tokens_processed": 15794176000, "tokens_processed_billions": 15.794176}
+{"timestamp": 1787648690.6297991, "elapsed_hours": 5.953811915781763, "session_id": 2, "event": "lambda_stats", "step": 120500, "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": 1787648716.356595, "elapsed_hours": 5.960958247913255, "session_id": 2, "event": "validation", "step": 120500, "epoch": 1, "val_loss": 1.7515667796134948, "val_perplexity": 5.7636259370471254, "is_best": false, "best_loss": 1.6233880519866943, "best_perplexity": 5.070239483980759, "train_loss": 1.216089129447937, "train_loss_avg": 1.6548359680175782}