Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -2427,3 +2427,31 @@
 {"timestamp": 1787644656.508547, "elapsed_hours": 4.83322267903222, "session_id": 2, "event": "train_step", "step": 115500, "epoch": 1, "loss": 1.8112013638019562, "loss_raw": 2.0083930492401123, "perplexity": 6.117792713883756, "muon_lr": 0.0021540379287008863, "adamw_lr": 6.462113786102658e-05, "tokens_per_sec": 874028.0480452776, "vram_gb": 2.785593032836914, "grad_norm": 0.09716796875, "tokens_processed": 15138816000, "tokens_processed_billions": 15.138816}
 {"timestamp": 1787644656.5103605, "elapsed_hours": 4.8332231828239225, "session_id": 2, "event": "lambda_stats", "step": 115500, "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": 1787644682.224807, "elapsed_hours": 4.840366084575653, "session_id": 2, "event": "validation", "step": 115500, "epoch": 1, "val_loss": 1.754755751490593, "val_perplexity": 5.782035315965061, "is_best": false, "best_loss": 1.6233880519866943, "best_perplexity": 5.070239483980759, "train_loss": 2.0083930492401123, "train_loss_avg": 1.8112013638019562}
+{"timestamp": 1787644690.23305, "elapsed_hours": 4.842590596609646, "session_id": 2, "event": "train_step", "step": 115510, "epoch": 1, "loss": 1.8079054820537568, "loss_raw": 1.7372936010360718, "perplexity": 6.097662384354343, "muon_lr": 0.0021535269882577705, "adamw_lr": 6.46058096477331e-05, "tokens_per_sec": 872405.0964174252, "vram_gb": 2.785593032836914, "grad_norm": 0.1015625, "tokens_processed": 15140126720, "tokens_processed_billions": 15.14012672}
+{"timestamp": 1787644697.7966225, "elapsed_hours": 4.844691588812404, "session_id": 2, "event": "train_step", "step": 115520, "epoch": 1, "loss": 1.7939344084262847, "loss_raw": 1.6941176652908325, "perplexity": 6.013063836564769, "muon_lr": 0.002152505396897139, "adamw_lr": 6.457516190691417e-05, "tokens_per_sec": 872100.5347351624, "vram_gb": 2.363645076751709, "grad_norm": 0.08935546875, "tokens_processed": 15141437440, "tokens_processed_billions": 15.14143744}
+{"timestamp": 1787644705.2743244, "elapsed_hours": 4.846768728362189, "session_id": 2, "event": "train_step", "step": 115530, "epoch": 1, "loss": 1.8168644714355469, "loss_raw": 1.830736756324768, "perplexity": 6.152536719167663, "muon_lr": 0.0021519947460090826, "adamw_lr": 6.455984238027247e-05, "tokens_per_sec": 871800.5524386999, "vram_gb": 2.785593032836914, "grad_norm": 0.09521484375, "tokens_processed": 15142748160, "tokens_processed_billions": 15.14274816}
+{"timestamp": 1787644712.7175786, "elapsed_hours": 4.848836298916075, "session_id": 2, "event": "train_step", "step": 115540, "epoch": 1, "loss": 1.7617408812046051, "loss_raw": 2.405949831008911, "perplexity": 5.822564970310946, "muon_lr": 0.002151484191668835, "adamw_lr": 6.454452575006504e-05, "tokens_per_sec": 871502.5526375511, "vram_gb": 2.785593032836914, "grad_norm": 0.09521484375, "tokens_processed": 15144058880, "tokens_processed_billions": 15.14405888}
+{"timestamp": 1787644720.191454, "elapsed_hours": 4.850912375317679, "session_id": 2, "event": "train_step", "step": 115550, "epoch": 1, "loss": 1.7627244234085082, "loss_raw": 1.129696249961853, "perplexity": 5.8282945258621845, "muon_lr": 0.0021509737338911176, "adamw_lr": 6.452921201673352e-05, "tokens_per_sec": 871203.2766026489, "vram_gb": 2.785593032836914, "grad_norm": 0.10107421875, "tokens_processed": 15145369600, "tokens_processed_billions": 15.1453696}
+{"timestamp": 1787644727.7566283, "elapsed_hours": 4.853013812700907, "session_id": 2, "event": "train_step", "step": 115560, "epoch": 1, "loss": 1.7121487712860108, "loss_raw": 1.918516755104065, "perplexity": 5.540854724646289, "muon_lr": 0.0021499531080821544, "adamw_lr": 6.449859324246462e-05, "tokens_per_sec": 870899.6820782142, "vram_gb": 2.363645076751709, "grad_norm": 0.087890625, "tokens_processed": 15146680320, "tokens_processed_billions": 15.14668032}
+{"timestamp": 1787644735.1986141, "elapsed_hours": 4.855081030925115, "session_id": 2, "event": "train_step", "step": 115570, "epoch": 1, "loss": 1.7340668725967407, "loss_raw": 1.4947311878204346, "perplexity": 5.663640439200207, "muon_lr": 0.0021494429400803414, "adamw_lr": 6.448328820241023e-05, "tokens_per_sec": 870602.5182300186, "vram_gb": 2.785593032836914, "grad_norm": 0.09912109375, "tokens_processed": 15147991040, "tokens_processed_billions": 15.14799104}
+{"timestamp": 1787644742.681927, "elapsed_hours": 4.857159729003906, "session_id": 2, "event": "train_step", "step": 115580, "epoch": 1, "loss": 1.6993629217147828, "loss_raw": 1.66020667552948, "perplexity": 5.4704611693257155, "muon_lr": 0.002148932868699924, "adamw_lr": 6.446798606099771e-05, "tokens_per_sec": 870303.5433766337, "vram_gb": 2.785593032836914, "grad_norm": 0.09326171875, "tokens_processed": 15149301760, "tokens_processed_billions": 15.14930176}
+{"timestamp": 1787644750.1320415, "elapsed_hours": 4.85922920525074, "session_id": 2, "event": "train_step", "step": 115590, "epoch": 1, "loss": 1.7186090421676636, "loss_raw": 2.4012365341186523, "perplexity": 5.576766020536264, "muon_lr": 0.0021484228939556097, "adamw_lr": 6.445268681866829e-05, "tokens_per_sec": 870006.4777409354, "vram_gb": 2.785593032836914, "grad_norm": 0.09228515625, "tokens_processed": 15150612480, "tokens_processed_billions": 15.15061248}
+{"timestamp": 1787644757.6869411, "elapsed_hours": 4.8613277884986665, "session_id": 2, "event": "train_step", "step": 115600, "epoch": 1, "loss": 1.6647966194152832, "loss_raw": 1.9539693593978882, "perplexity": 5.2845983557545955, "muon_lr": 0.0021474032344341197, "adamw_lr": 6.442209703302357e-05, "tokens_per_sec": 869704.4376242661, "vram_gb": 2.363645076751709, "grad_norm": 0.09423828125, "tokens_processed": 15151923200, "tokens_processed_billions": 15.1519232}
+{"timestamp": 1787644757.689304, "elapsed_hours": 4.861328444944488, "session_id": 2, "event": "lambda_stats", "step": 115600, "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": 1787644765.1953294, "elapsed_hours": 4.863413451976246, "session_id": 2, "event": "train_step", "step": 115610, "epoch": 1, "loss": 1.6998719048500062, "loss_raw": 2.29616641998291, "perplexity": 5.473246250522351, "muon_lr": 0.002146893549686348, "adamw_lr": 6.440680649059044e-05, "tokens_per_sec": 869404.9834674462, "vram_gb": 2.785593032836914, "grad_norm": 0.09521484375, "tokens_processed": 15153233920, "tokens_processed_billions": 15.15323392}
+{"timestamp": 1787644772.6440296, "elapsed_hours": 4.8654825352960165, "session_id": 2, "event": "train_step", "step": 115620, "epoch": 1, "loss": 1.7100914573669435, "loss_raw": 1.799227237701416, "perplexity": 5.5294671650067455, "muon_lr": 0.002146383961633488, "adamw_lr": 6.439151884900463e-05, "tokens_per_sec": 869108.7597362663, "vram_gb": 2.785593032836914, "grad_norm": 0.09521484375, "tokens_processed": 15154544640, "tokens_processed_billions": 15.15454464}
+{"timestamp": 1787644780.1580224, "elapsed_hours": 4.867569755448235, "session_id": 2, "event": "train_step", "step": 115630, "epoch": 1, "loss": 1.7593706774711608, "loss_raw": 1.3906559944152832, "perplexity": 5.808780647360159, "muon_lr": 0.0021458744702902385, "adamw_lr": 6.437623410870715e-05, "tokens_per_sec": 868809.5372078605, "vram_gb": 2.785593032836914, "grad_norm": 0.10595703125, "tokens_processed": 15155855360, "tokens_processed_billions": 15.15585536}
+{"timestamp": 1787644787.7074857, "elapsed_hours": 4.869666828645601, "session_id": 2, "event": "train_step", "step": 115640, "epoch": 1, "loss": 1.7329646682739257, "loss_raw": 2.0646743774414062, "perplexity": 5.657401389210813, "muon_lr": 0.00214485577779133, "adamw_lr": 6.434567333373987e-05, "tokens_per_sec": 868508.8079426799, "vram_gb": 2.363645076751709, "grad_norm": 0.09375, "tokens_processed": 15157166080, "tokens_processed_billions": 15.15716608}
+{"timestamp": 1787644795.154019, "elapsed_hours": 4.871735310223367, "session_id": 2, "event": "train_step", "step": 115650, "epoch": 1, "loss": 1.7398009133338928, "loss_raw": 1.3992714881896973, "perplexity": 5.69620927046534, "muon_lr": 0.002144346576665049, "adamw_lr": 6.433039729995146e-05, "tokens_per_sec": 868213.4553319701, "vram_gb": 2.785593032836914, "grad_norm": 0.09521484375, "tokens_processed": 15158476800, "tokens_processed_billions": 15.1584768}
+{"timestamp": 1787644802.6412375, "elapsed_hours": 4.873815093172921, "session_id": 2, "event": "train_step", "step": 115660, "epoch": 1, "loss": 1.7097454249858857, "loss_raw": 2.350705146789551, "perplexity": 5.527554121324276, "muon_lr": 0.0021438374723071283, "adamw_lr": 6.431512416921383e-05, "tokens_per_sec": 867916.3350174167, "vram_gb": 2.785593032836914, "grad_norm": 0.087890625, "tokens_processed": 15159787520, "tokens_processed_billions": 15.15978752}
+{"timestamp": 1787644810.0998964, "elapsed_hours": 4.8758869427442555, "session_id": 2, "event": "train_step", "step": 115670, "epoch": 1, "loss": 1.692825621366501, "loss_raw": 1.2634519338607788, "perplexity": 5.434815760934535, "muon_lr": 0.0021433284647322517, "adamw_lr": 6.429985394196754e-05, "tokens_per_sec": 867620.8847776061, "vram_gb": 2.785593032836914, "grad_norm": 0.09423828125, "tokens_processed": 15161098240, "tokens_processed_billions": 15.16109824}
+{"timestamp": 1787644817.640076, "elapsed_hours": 4.87798143711355, "session_id": 2, "event": "train_step", "step": 115680, "epoch": 1, "loss": 1.577529762983322, "loss_raw": 1.4265869855880737, "perplexity": 4.842977720131128, "muon_lr": 0.002142310739990336, "adamw_lr": 6.426932219971007e-05, "tokens_per_sec": 867321.6410808486, "vram_gb": 2.363645076751709, "grad_norm": 0.087890625, "tokens_processed": 15162408960, "tokens_processed_billions": 15.16240896}
+{"timestamp": 1787644825.139363, "elapsed_hours": 4.88006457236078, "session_id": 2, "event": "train_step", "step": 115690, "epoch": 1, "loss": 1.5775987637042999, "loss_raw": 1.601373553276062, "perplexity": 4.843311900614712, "muon_lr": 0.0021418020228526477, "adamw_lr": 6.425406068557942e-05, "tokens_per_sec": 867024.6870392987, "vram_gb": 2.785593032836914, "grad_norm": 0.09716796875, "tokens_processed": 15163719680, "tokens_processed_billions": 15.16371968}
+{"timestamp": 1787644832.583829, "elapsed_hours": 4.882132479482227, "session_id": 2, "event": "train_step", "step": 115700, "epoch": 1, "loss": 1.6366899049282073, "loss_raw": 1.7193864583969116, "perplexity": 5.138133621801261, "muon_lr": 0.0021412934025566975, "adamw_lr": 6.42388020767009e-05, "tokens_per_sec": 866730.7022951349, "vram_gb": 2.785593032836914, "grad_norm": 0.09619140625, "tokens_processed": 15165030400, "tokens_processed_billions": 15.1650304}
+{"timestamp": 1787644832.5856314, "elapsed_hours": 4.882132980293698, "session_id": 2, "event": "lambda_stats", "step": 115700, "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": 1787644840.03894, "elapsed_hours": 4.884203343656328, "session_id": 2, "event": "train_step", "step": 115710, "epoch": 1, "loss": 1.6427416276931763, "loss_raw": 2.041269063949585, "perplexity": 5.169322459924027, "muon_lr": 0.0021407848791171556, "adamw_lr": 6.422354637351465e-05, "tokens_per_sec": 866436.4386133751, "vram_gb": 2.785593032836914, "grad_norm": 0.09765625, "tokens_processed": 15166341120, "tokens_processed_billions": 15.16634112}
+{"timestamp": 1787644847.5881758, "elapsed_hours": 4.886300353738997, "session_id": 2, "event": "train_step", "step": 115720, "epoch": 1, "loss": 1.6148888325691224, "loss_raw": 1.7818505764007568, "perplexity": 5.027329017152222, "muon_lr": 0.002139768122865947, "adamw_lr": 6.419304368597839e-05, "tokens_per_sec": 866137.769319938, "vram_gb": 2.363645076751709, "grad_norm": 0.09130859375, "tokens_processed": 15167651840, "tokens_processed_billions": 15.16765184}
+{"timestamp": 1787644855.0262709, "elapsed_hours": 4.88836649119854, "session_id": 2, "event": "train_step", "step": 115730, "epoch": 1, "loss": 1.628171330690384, "loss_raw": 1.0425114631652832, "perplexity": 5.094549948038154, "muon_lr": 0.0021392598900836024, "adamw_lr": 6.417779670250805e-05, "tokens_per_sec": 865844.8553282869, "vram_gb": 2.785593032836914, "grad_norm": 0.09033203125, "tokens_processed": 15168962560, "tokens_processed_billions": 15.16896256}
+{"timestamp": 1787644862.478778, "elapsed_hours": 4.890436631970935, "session_id": 2, "event": "train_step", "step": 115740, "epoch": 1, "loss": 1.6908420360088348, "loss_raw": 2.2390024662017822, "perplexity": 5.424046024846164, "muon_lr": 0.0021387517542163044, "adamw_lr": 6.416255262648913e-05, "tokens_per_sec": 865551.4736189557, "vram_gb": 2.785593032836914, "grad_norm": 0.09814453125, "tokens_processed": 15170273280, "tokens_processed_billions": 15.17027328}
+{"timestamp": 1787644869.9271443, "elapsed_hours": 4.892505622638597, "session_id": 2, "event": "train_step", "step": 115750, "epoch": 1, "loss": 1.604896878004074, "loss_raw": 1.9922435283660889, "perplexity": 4.977346302388744, "muon_lr": 0.0021382437152787097, "adamw_lr": 6.414731145836128e-05, "tokens_per_sec": 865258.5475106116, "vram_gb": 2.785593032836914, "grad_norm": 0.095703125, "tokens_processed": 15171584000, "tokens_processed_billions": 15.171584}
+{"timestamp": 1787644877.4669154, "elapsed_hours": 4.8946000035603845, "session_id": 2, "event": "train_step", "step": 115760, "epoch": 1, "loss": 1.6317022383213042, "loss_raw": 1.86037015914917, "perplexity": 5.112570128400036, "muon_lr": 0.0021372279282512235, "adamw_lr": 6.411683784753669e-05, "tokens_per_sec": 864961.3645233581, "vram_gb": 2.363645076751709, "grad_norm": 0.09765625, "tokens_processed": 15172894720, "tokens_processed_billions": 15.17289472}