Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -5052,3 +5052,34 @@
 {"timestamp": 1787546236.1768198, "elapsed_hours": 5.44785786456532, "session_id": 2, "event": "train_step", "step": 44710, "epoch": 1, "loss": 1.7858576107025146, "loss_raw": 2.2908401489257812, "perplexity": 5.964693139378736, "muon_lr": 0.008274430870087083, "adamw_lr": 0.00024823292610261244, "tokens_per_sec": 149487.632359739, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 2930114560, "tokens_processed_billions": 2.93011456}
 {"timestamp": 1787546240.2647903, "elapsed_hours": 5.448993411925104, "session_id": 2, "event": "train_step", "step": 44720, "epoch": 1, "loss": 1.856523985862732, "loss_raw": 1.7565408945083618, "perplexity": 6.401446534630478, "muon_lr": 0.008273227410330493, "adamw_lr": 0.0002481968223099148, "tokens_per_sec": 149489.88891330487, "vram_gb": 27.299458026885986, "grad_norm": 0.1494140625, "tokens_processed": 2930769920, "tokens_processed_billions": 2.93076992}
 {"timestamp": 1787546244.1915607, "elapsed_hours": 5.450084181427956, "session_id": 2, "event": "train_step", "step": 44730, "epoch": 1, "loss": 1.845556037425995, "loss_raw": 0.8913135528564453, "perplexity": 6.331619429155718, "muon_lr": 0.008272625560484467, "adamw_lr": 0.00024817876681453396, "tokens_per_sec": 149493.37509771384, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2931425280, "tokens_processed_billions": 2.93142528}
+{"timestamp": 1787546248.9284506, "elapsed_hours": 5.451399984161059, "session_id": 2, "event": "train_step", "step": 44740, "epoch": 1, "loss": 1.881949087381363, "loss_raw": 1.8373639583587646, "perplexity": 6.566290672104415, "muon_lr": 0.008272023630683096, "adamw_lr": 0.0002481607089204928, "tokens_per_sec": 149490.68438473303, "vram_gb": 27.710740089416504, "grad_norm": 0.13671875, "tokens_processed": 2932080640, "tokens_processed_billions": 2.93208064}
+{"timestamp": 1787546252.854952, "elapsed_hours": 5.452490679158105, "session_id": 2, "event": "train_step", "step": 44750, "epoch": 1, "loss": 1.923941205739975, "loss_raw": 2.439741611480713, "perplexity": 6.847894314950382, "muon_lr": 0.008271421620943735, "adamw_lr": 0.000248142648628312, "tokens_per_sec": 149494.17034017187, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 2932736000, "tokens_processed_billions": 2.932736}
+{"timestamp": 1787546256.9532769, "elapsed_hours": 5.453629102574454, "session_id": 2, "event": "train_step", "step": 44760, "epoch": 1, "loss": 1.8911809098720551, "loss_raw": 2.5168373584747314, "perplexity": 6.62719017619676, "muon_lr": 0.00827021736172049, "adamw_lr": 0.00024810652085161466, "tokens_per_sec": 149496.34509605772, "vram_gb": 27.299458026885986, "grad_norm": 0.14453125, "tokens_processed": 2933391360, "tokens_processed_billions": 2.93339136}
+{"timestamp": 1787546260.870813, "elapsed_hours": 5.454717307024532, "session_id": 2, "event": "train_step", "step": 44770, "epoch": 1, "loss": 1.8813920438289642, "loss_raw": 1.3766618967056274, "perplexity": 6.56263398078506, "muon_lr": 0.008269615112271336, "adamw_lr": 0.00024808845336814, "tokens_per_sec": 149499.89736971474, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 2934046720, "tokens_processed_billions": 2.93404672}
+{"timestamp": 1787546264.9277697, "elapsed_hours": 5.455844239592552, "session_id": 2, "event": "train_step", "step": 44780, "epoch": 1, "loss": 1.951904228925705, "loss_raw": 2.1822259426116943, "perplexity": 7.042084562201727, "muon_lr": 0.008269012782953649, "adamw_lr": 0.0002480703834886094, "tokens_per_sec": 149502.38557040814, "vram_gb": 27.710740089416504, "grad_norm": 0.162109375, "tokens_processed": 2934702080, "tokens_processed_billions": 2.93470208}
+{"timestamp": 1787546268.9024196, "elapsed_hours": 5.4569483088784745, "session_id": 2, "event": "train_step", "step": 44790, "epoch": 1, "loss": 1.9918340277671813, "loss_raw": 2.1317927837371826, "perplexity": 7.32896296536277, "muon_lr": 0.008268410373784797, "adamw_lr": 0.00024805231121354383, "tokens_per_sec": 149505.50014206357, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2935357440, "tokens_processed_billions": 2.93535744}
+{"timestamp": 1787546272.9855926, "elapsed_hours": 5.458082523677084, "session_id": 2, "event": "train_step", "step": 44800, "epoch": 1, "loss": 1.9465919303894044, "loss_raw": 1.6213064193725586, "perplexity": 7.004774096598685, "muon_lr": 0.008267205315963096, "adamw_lr": 0.00024801615947889276, "tokens_per_sec": 149507.78573797093, "vram_gb": 27.299458026885986, "grad_norm": 0.1474609375, "tokens_processed": 2936012800, "tokens_processed_billions": 2.9360128}
+{"timestamp": 1787546272.9895585, "elapsed_hours": 5.458083625502057, "session_id": 2, "event": "lambda_stats", "step": 44800, "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": 1787546277.034163, "elapsed_hours": 5.45920712656445, "session_id": 2, "event": "train_step", "step": 44810, "epoch": 1, "loss": 1.9701874089241027, "loss_raw": 3.3050994873046875, "perplexity": 7.172020463044875, "muon_lr": 0.008266602667344995, "adamw_lr": 0.0002479980800203498, "tokens_per_sec": 149510.3351508677, "vram_gb": 27.710740089416504, "grad_norm": 0.1328125, "tokens_processed": 2936668160, "tokens_processed_billions": 2.93666816}
+{"timestamp": 1787546280.996214, "elapsed_hours": 5.460307696262995, "session_id": 2, "event": "train_step", "step": 44820, "epoch": 1, "loss": 1.9740171766281127, "loss_raw": 2.4428603649139404, "perplexity": 7.199540299039913, "muon_lr": 0.00826599993894523, "adamw_lr": 0.00024797999816835684, "tokens_per_sec": 149513.5420405349, "vram_gb": 27.710740089416504, "grad_norm": 0.16015625, "tokens_processed": 2937323520, "tokens_processed_billions": 2.93732352}
+{"timestamp": 1787546284.9720519, "elapsed_hours": 5.461412095692423, "session_id": 2, "event": "train_step", "step": 44830, "epoch": 1, "loss": 1.9614549899101257, "loss_raw": 2.031075954437256, "perplexity": 7.109664032673227, "muon_lr": 0.008265397130781189, "adamw_lr": 0.00024796191392343557, "tokens_per_sec": 149516.64256200497, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 2937978880, "tokens_processed_billions": 2.93797888}
+{"timestamp": 1787546289.1639757, "elapsed_hours": 5.46257651898596, "session_id": 2, "event": "train_step", "step": 44840, "epoch": 1, "loss": 1.9169518375396728, "loss_raw": 2.6035685539245605, "perplexity": 6.800198735335802, "muon_lr": 0.008264191275229794, "adamw_lr": 0.0002479257382568938, "tokens_per_sec": 149518.0963556916, "vram_gb": 27.299458026885986, "grad_norm": 0.1494140625, "tokens_processed": 2938634240, "tokens_processed_billions": 2.93863424}
+{"timestamp": 1787546293.1277297, "elapsed_hours": 5.463677561746703, "session_id": 2, "event": "train_step", "step": 44850, "epoch": 1, "loss": 1.9350762581825256, "loss_raw": 2.0197248458862305, "perplexity": 6.924572090526912, "muon_lr": 0.00826358822787722, "adamw_lr": 0.0002479076468363165, "tokens_per_sec": 149521.28610876028, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 2939289600, "tokens_processed_billions": 2.9392896}
+{"timestamp": 1787546297.0799198, "elapsed_hours": 5.464775392479367, "session_id": 2, "event": "train_step", "step": 44860, "epoch": 1, "loss": 1.9829853367805481, "loss_raw": 1.9221636056900024, "perplexity": 7.264397318843696, "muon_lr": 0.008262985100829907, "adamw_lr": 0.0002478895530248972, "tokens_per_sec": 149524.56336769005, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 2939944960, "tokens_processed_billions": 2.93994496}
+{"timestamp": 1787546301.0375204, "elapsed_hours": 5.465874725845125, "session_id": 2, "event": "train_step", "step": 44870, "epoch": 1, "loss": 1.955326066017151, "loss_raw": 2.0221405029296875, "perplexity": 7.066222703198134, "muon_lr": 0.008262381894105258, "adamw_lr": 0.00024787145682315774, "tokens_per_sec": 149527.7976677556, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 2940600320, "tokens_processed_billions": 2.94060032}
+{"timestamp": 1787546305.2492278, "elapsed_hours": 5.4670446445544565, "session_id": 2, "event": "train_step", "step": 44880, "epoch": 1, "loss": 1.9114165306091309, "loss_raw": 1.3903738260269165, "perplexity": 6.762661533956268, "muon_lr": 0.008261175241693518, "adamw_lr": 0.0002478352572508055, "tokens_per_sec": 149529.09896276737, "vram_gb": 27.299458026885986, "grad_norm": 0.1455078125, "tokens_processed": 2941255680, "tokens_processed_billions": 2.94125568}
+{"timestamp": 1787546309.2073243, "elapsed_hours": 5.468144115871853, "session_id": 2, "event": "train_step", "step": 44890, "epoch": 1, "loss": 1.9157476603984833, "loss_raw": 2.01479172706604, "perplexity": 6.792015019773665, "muon_lr": 0.008260571796041227, "adamw_lr": 0.00024781715388123675, "tokens_per_sec": 149532.32672370053, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2941911040, "tokens_processed_billions": 2.94191104}
+{"timestamp": 1787546313.1570292, "elapsed_hours": 5.4692412560515935, "session_id": 2, "event": "train_step", "step": 44900, "epoch": 1, "loss": 1.901854751110077, "loss_raw": 1.0895761251449585, "perplexity": 6.6983066196095224, "muon_lr": 0.008259968270781185, "adamw_lr": 0.00024779904812343546, "tokens_per_sec": 149535.61702202304, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 2942566400, "tokens_processed_billions": 2.9425664}
+{"timestamp": 1787546313.1629326, "elapsed_hours": 5.4692428962389625, "session_id": 2, "event": "lambda_stats", "step": 44900, "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": 1787546317.1430435, "elapsed_hours": 5.470348482198186, "session_id": 2, "event": "train_step", "step": 44910, "epoch": 1, "loss": 1.911320037841797, "loss_raw": 2.904128313064575, "perplexity": 6.762009017512379, "muon_lr": 0.0082593646659308, "adamw_lr": 0.00024778093997792396, "tokens_per_sec": 149538.6308596655, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 2943221760, "tokens_processed_billions": 2.94322176}
+{"timestamp": 1787546321.3107536, "elapsed_hours": 5.4715061794386965, "session_id": 2, "event": "train_step", "step": 44920, "epoch": 1, "loss": 1.9402548801898956, "loss_raw": 2.0426111221313477, "perplexity": 6.960524844459149, "muon_lr": 0.008258157217528633, "adamw_lr": 0.00024774471652585893, "tokens_per_sec": 149540.26238637386, "vram_gb": 27.299458026885986, "grad_norm": 0.1455078125, "tokens_processed": 2943877120, "tokens_processed_billions": 2.94387712}
+{"timestamp": 1787546325.3340938, "elapsed_hours": 5.472623774078158, "session_id": 2, "event": "train_step", "step": 44930, "epoch": 1, "loss": 1.955195631980896, "loss_raw": 2.3239715099334717, "perplexity": 7.065301087356224, "muon_lr": 0.008257553374011666, "adamw_lr": 0.00024772660122034996, "tokens_per_sec": 149542.99041901086, "vram_gb": 27.710740089416504, "grad_norm": 0.1337890625, "tokens_processed": 2944532480, "tokens_processed_billions": 2.94453248}
+{"timestamp": 1787546329.3930883, "elapsed_hours": 5.473751272559166, "session_id": 2, "event": "train_step", "step": 44940, "epoch": 1, "loss": 2.0522036015987397, "loss_raw": 1.3666541576385498, "perplexity": 7.785037339384523, "muon_lr": 0.008256949450973997, "adamw_lr": 0.0002477084835292199, "tokens_per_sec": 149545.44594738685, "vram_gb": 27.710740089416504, "grad_norm": 0.17578125, "tokens_processed": 2945187840, "tokens_processed_billions": 2.94518784}
+{"timestamp": 1787546333.4408038, "elapsed_hours": 5.474875637955136, "session_id": 2, "event": "train_step", "step": 44950, "epoch": 1, "loss": 2.091613965034485, "loss_raw": 1.1658915281295776, "perplexity": 8.097974470483969, "muon_lr": 0.00825634544843304, "adamw_lr": 0.0002476903634529912, "tokens_per_sec": 149547.9866150871, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 2945843200, "tokens_processed_billions": 2.9458432}
+{"timestamp": 1787546337.5711744, "elapsed_hours": 5.476022963126501, "session_id": 2, "event": "train_step", "step": 44960, "epoch": 1, "loss": 2.0269952809810636, "loss_raw": 1.6484359502792358, "perplexity": 7.591242505552883, "muon_lr": 0.008255137204910935, "adamw_lr": 0.00024765411614732796, "tokens_per_sec": 149549.89846104506, "vram_gb": 27.299458026885986, "grad_norm": 0.150390625, "tokens_processed": 2946498560, "tokens_processed_billions": 2.94649856}
+{"timestamp": 1787546341.593894, "elapsed_hours": 5.477140385111173, "session_id": 2, "event": "train_step", "step": 44970, "epoch": 1, "loss": 1.9666379594802856, "loss_raw": 2.486457586288452, "perplexity": 7.14660886427303, "muon_lr": 0.00825453296396463, "adamw_lr": 0.0002476359889189389, "tokens_per_sec": 149552.62664499544, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 2947153920, "tokens_processed_billions": 2.94715392}
+{"timestamp": 1787546345.611255, "elapsed_hours": 5.4782563187678655, "session_id": 2, "event": "train_step", "step": 44980, "epoch": 1, "loss": 1.9598997366428375, "loss_raw": 1.6728460788726807, "perplexity": 7.09861529847414, "muon_lr": 0.008253928643584721, "adamw_lr": 0.00024761785930754157, "tokens_per_sec": 149555.3946101917, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 2947809280, "tokens_processed_billions": 2.94780928}
+{"timestamp": 1787546349.6850333, "elapsed_hours": 5.479387923942672, "session_id": 2, "event": "train_step", "step": 44990, "epoch": 1, "loss": 1.8794603908061982, "loss_raw": 2.97861385345459, "perplexity": 6.54996948466278, "muon_lr": 0.008253324243788638, "adamw_lr": 0.00024759972731365905, "tokens_per_sec": 149557.73358636288, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 2948464640, "tokens_processed_billions": 2.94846464}
+{"timestamp": 1787546353.7984495, "elapsed_hours": 5.480530539486144, "session_id": 2, "event": "train_step", "step": 45000, "epoch": 1, "loss": 1.8696460795402527, "loss_raw": 2.1941282749176025, "perplexity": 6.4860004647551674, "muon_lr": 0.008252115206017662, "adamw_lr": 0.0002475634561805298, "tokens_per_sec": 149559.76950524547, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 2949120000, "tokens_processed_billions": 2.94912}
+{"timestamp": 1787546353.801489, "elapsed_hours": 5.480531384017732, "session_id": 2, "event": "lambda_stats", "step": 45000, "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": 1787546367.2483637, "elapsed_hours": 5.484266627497143, "session_id": 2, "event": "validation", "step": 45000, "epoch": 1, "val_loss": 1.9518783861398696, "val_perplexity": 7.041902577470059, "is_best": true, "best_loss": 1.9518783861398696, "best_perplexity": 7.041902577470059, "train_loss": 2.1941282749176025, "train_loss_avg": 1.8696460795402527}