Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -2625,3 +2625,31 @@
 {"timestamp": 1787646072.1639662, "elapsed_hours": 5.226460295385785, "session_id": 2, "event": "train_step", "step": 117240, "epoch": 1, "loss": 1.7094597625732422, "loss_raw": 2.2337796688079834, "perplexity": 5.5259753323890255, "muon_lr": 0.0020449603436846816, "adamw_lr": 6.134881031054044e-05, "tokens_per_sec": 820161.9193646626, "vram_gb": 2.363645076751709, "grad_norm": 0.095703125, "tokens_processed": 15366881280, "tokens_processed_billions": 15.36688128}
 {"timestamp": 1787646079.5967004, "elapsed_hours": 5.228524943855073, "session_id": 2, "event": "train_step", "step": 117250, "epoch": 1, "loss": 1.6792448580265045, "loss_raw": 1.9051164388656616, "perplexity": 5.3615057440377445, "muon_lr": 0.002044470784694779, "adamw_lr": 6.133412354084336e-05, "tokens_per_sec": 819906.6195402044, "vram_gb": 2.785593032836914, "grad_norm": 0.09228515625, "tokens_processed": 15368192000, "tokens_processed_billions": 15.368192}
 {"timestamp": 1787646087.042285, "elapsed_hours": 5.230593161715402, "session_id": 2, "event": "train_step", "step": 117260, "epoch": 1, "loss": 1.7091453111171722, "loss_raw": 1.477288842201233, "perplexity": 5.524237954574357, "muon_lr": 0.0020439813253533947, "adamw_lr": 6.131943976060184e-05, "tokens_per_sec": 819650.9589415143, "vram_gb": 2.785593032836914, "grad_norm": 0.10009765625, "tokens_processed": 15369502720, "tokens_processed_billions": 15.36950272}
+{"timestamp": 1787646095.2433074, "elapsed_hours": 5.232871223555671, "session_id": 2, "event": "train_step", "step": 117270, "epoch": 1, "loss": 1.6983320677280427, "loss_raw": 1.7079663276672363, "perplexity": 5.464824828241475, "muon_lr": 0.0020434919656746485, "adamw_lr": 6.130475897023945e-05, "tokens_per_sec": 819362.5053675891, "vram_gb": 2.785593032836914, "grad_norm": 0.09716796875, "tokens_processed": 15370813440, "tokens_processed_billions": 15.37081344}
+{"timestamp": 1787646102.7731738, "elapsed_hours": 5.2349628530608285, "session_id": 2, "event": "train_step", "step": 117280, "epoch": 1, "loss": 1.7150748336315156, "loss_raw": 1.02363920211792, "perplexity": 5.557091354128444, "muon_lr": 0.0020425135453615057, "adamw_lr": 6.127540636084516e-05, "tokens_per_sec": 819103.5916649284, "vram_gb": 2.363645076751709, "grad_norm": 0.0908203125, "tokens_processed": 15372124160, "tokens_processed_billions": 15.37212416}
+{"timestamp": 1787646110.215326, "elapsed_hours": 5.2370301175779765, "session_id": 2, "event": "train_step", "step": 117290, "epoch": 1, "loss": 1.6905485951900483, "loss_raw": 0.9425482153892517, "perplexity": 5.422454621842217, "muon_lr": 0.0020420244847553256, "adamw_lr": 6.126073454265976e-05, "tokens_per_sec": 818848.7175708464, "vram_gb": 2.785593032836914, "grad_norm": 0.0927734375, "tokens_processed": 15373434880, "tokens_processed_billions": 15.37343488}
+{"timestamp": 1787646117.664716, "elapsed_hours": 5.239099392626021, "session_id": 2, "event": "train_step", "step": 117300, "epoch": 1, "loss": 1.6600782036781312, "loss_raw": 1.1753548383712769, "perplexity": 5.259722157982276, "muon_lr": 0.0020415355238682102, "adamw_lr": 6.12460657160463e-05, "tokens_per_sec": 818593.7283992436, "vram_gb": 2.785593032836914, "grad_norm": 0.0859375, "tokens_processed": 15374745600, "tokens_processed_billions": 15.3747456}
+{"timestamp": 1787646117.666417, "elapsed_hours": 5.239099865092172, "session_id": 2, "event": "lambda_stats", "step": 117300, "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": 1787646125.1110706, "elapsed_hours": 5.2411678245332505, "session_id": 2, "event": "train_step", "step": 117310, "epoch": 1, "loss": 1.6282286834716797, "loss_raw": 1.890068531036377, "perplexity": 5.094842143026142, "muon_lr": 0.0020410466627142623, "adamw_lr": 6.123139988142785e-05, "tokens_per_sec": 818339.0704376774, "vram_gb": 2.785593032836914, "grad_norm": 0.109375, "tokens_processed": 15376056320, "tokens_processed_billions": 15.37605632}
+{"timestamp": 1787646132.6606896, "elapsed_hours": 5.243264940844642, "session_id": 2, "event": "train_step", "step": 117320, "epoch": 1, "loss": 1.5853355991840363, "loss_raw": 1.8177740573883057, "perplexity": 4.880929139518293, "muon_lr": 0.0020400692396622504, "adamw_lr": 6.120207718986751e-05, "tokens_per_sec": 818080.1216098401, "vram_gb": 2.363645076751709, "grad_norm": 0.087890625, "tokens_processed": 15377367040, "tokens_processed_billions": 15.37736704}
+{"timestamp": 1787646140.1000054, "elapsed_hours": 5.245331417454613, "session_id": 2, "event": "train_step", "step": 117330, "epoch": 1, "loss": 1.5677321922779084, "loss_raw": 1.5851221084594727, "perplexity": 4.795759990725893, "muon_lr": 0.0020395806777923748, "adamw_lr": 6.118742033377123e-05, "tokens_per_sec": 817826.1788584157, "vram_gb": 2.785593032836914, "grad_norm": 0.09423828125, "tokens_processed": 15378677760, "tokens_processed_billions": 15.37867776}
+{"timestamp": 1787646147.542621, "elapsed_hours": 5.247398810651568, "session_id": 2, "event": "train_step", "step": 117340, "epoch": 1, "loss": 1.5900972008705139, "loss_raw": 1.0766265392303467, "perplexity": 4.904225600157321, "muon_lr": 0.002039092215712036, "adamw_lr": 6.117276647136107e-05, "tokens_per_sec": 817572.2931849122, "vram_gb": 2.785593032836914, "grad_norm": 0.09375, "tokens_processed": 15379988480, "tokens_processed_billions": 15.37998848}
+{"timestamp": 1787646154.9846563, "elapsed_hours": 5.24946604265107, "session_id": 2, "event": "train_step", "step": 117350, "epoch": 1, "loss": 1.6047403299808503, "loss_raw": 2.0962655544281006, "perplexity": 4.976567169651641, "muon_lr": 0.002038603853435324, "adamw_lr": 6.115811560305972e-05, "tokens_per_sec": 817318.632812828, "vram_gb": 2.785593032836914, "grad_norm": 0.09033203125, "tokens_processed": 15381299200, "tokens_processed_billions": 15.3812992}
+{"timestamp": 1787646162.5250154, "elapsed_hours": 5.2515605868895845, "session_id": 2, "event": "train_step", "step": 117360, "epoch": 1, "loss": 1.6395809400081633, "loss_raw": 2.1592447757720947, "perplexity": 5.153009639530694, "muon_lr": 0.002037627428349103, "adamw_lr": 6.112882285047308e-05, "tokens_per_sec": 817060.9073225866, "vram_gb": 2.363645076751709, "grad_norm": 0.10986328125, "tokens_processed": 15382609920, "tokens_processed_billions": 15.38260992}
+{"timestamp": 1787646169.9748628, "elapsed_hours": 5.25362998896175, "session_id": 2, "event": "train_step", "step": 117370, "epoch": 1, "loss": 1.6821469640731812, "loss_raw": 1.7595700025558472, "perplexity": 5.377088002022937, "muon_lr": 0.0020371393655677526, "adamw_lr": 6.111418096703256e-05, "tokens_per_sec": 816807.311220551, "vram_gb": 2.785593032836914, "grad_norm": 0.0947265625, "tokens_processed": 15383920640, "tokens_processed_billions": 15.38392064}
+{"timestamp": 1787646177.4547844, "elapsed_hours": 5.255707745022244, "session_id": 2, "event": "train_step", "step": 117380, "epoch": 1, "loss": 1.7960092043876648, "loss_raw": 1.57406747341156, "perplexity": 6.025552668537326, "muon_lr": 0.002036651402646341, "adamw_lr": 6.109954207939022e-05, "tokens_per_sec": 816552.6135791221, "vram_gb": 2.785593032836914, "grad_norm": 0.10498046875, "tokens_processed": 15385231360, "tokens_processed_billions": 15.38523136}
+{"timestamp": 1787646184.899542, "elapsed_hours": 5.2577757332060076, "session_id": 2, "event": "train_step", "step": 117390, "epoch": 1, "loss": 1.810953506231308, "loss_raw": 2.0403316020965576, "perplexity": 6.116276560546767, "muon_lr": 0.002036163539598942, "adamw_lr": 6.108490618796825e-05, "tokens_per_sec": 816299.6419016155, "vram_gb": 2.785593032836914, "grad_norm": 0.09912109375, "tokens_processed": 15386542080, "tokens_processed_billions": 15.38654208}
+{"timestamp": 1787646192.433647, "elapsed_hours": 5.259868540101581, "session_id": 2, "event": "train_step", "step": 117400, "epoch": 1, "loss": 1.8197317349910735, "loss_raw": 1.8107258081436157, "perplexity": 6.170202978283882, "muon_lr": 0.00203518811318245, "adamw_lr": 6.10556433954735e-05, "tokens_per_sec": 816042.999802127, "vram_gb": 2.363645076751709, "grad_norm": 0.09814453125, "tokens_processed": 15387852800, "tokens_processed_billions": 15.3878528}
+{"timestamp": 1787646192.4355984, "elapsed_hours": 5.259869082172711, "session_id": 2, "event": "lambda_stats", "step": 117400, "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": 1787646199.8743598, "elapsed_hours": 5.26193540473779, "session_id": 2, "event": "train_step", "step": 117410, "epoch": 1, "loss": 1.759298758506775, "loss_raw": 2.0237905979156494, "perplexity": 5.8083629008937825, "muon_lr": 0.002034700549841487, "adamw_lr": 6.10410164952446e-05, "tokens_per_sec": 815790.6059401792, "vram_gb": 2.785593032836914, "grad_norm": 0.09326171875, "tokens_processed": 15389163520, "tokens_processed_billions": 15.38916352}
+{"timestamp": 1787646207.3219218, "elapsed_hours": 5.264004172020488, "session_id": 2, "event": "train_step", "step": 117420, "epoch": 1, "loss": 1.7694103610515595, "loss_raw": 1.8794567584991455, "perplexity": 5.867392697951385, "muon_lr": 0.002034213086430793, "adamw_lr": 6.1026392592923795e-05, "tokens_per_sec": 815538.1132988876, "vram_gb": 2.785593032836914, "grad_norm": 0.09375, "tokens_processed": 15390474240, "tokens_processed_billions": 15.39047424}
+{"timestamp": 1787646214.7997077, "elapsed_hours": 5.266081334749858, "session_id": 2, "event": "train_step", "step": 117430, "epoch": 1, "loss": 1.6971599709987641, "loss_raw": 1.6209439039230347, "perplexity": 5.458423277285558, "muon_lr": 0.002033725722964424, "adamw_lr": 6.1011771688932714e-05, "tokens_per_sec": 815284.5161224661, "vram_gb": 2.785593032836914, "grad_norm": 0.0986328125, "tokens_processed": 15391784960, "tokens_processed_billions": 15.39178496}
+{"timestamp": 1787646222.3362954, "elapsed_hours": 5.268174831271171, "session_id": 2, "event": "train_step", "step": 117440, "epoch": 1, "loss": 1.7121174788475038, "loss_raw": 2.0877654552459717, "perplexity": 5.540681340503363, "muon_lr": 0.0020327512959208826, "adamw_lr": 6.098253887762646e-05, "tokens_per_sec": 815028.5773153837, "vram_gb": 2.363645076751709, "grad_norm": 0.10302734375, "tokens_processed": 15393095680, "tokens_processed_billions": 15.39309568}
+{"timestamp": 1787646229.768859, "elapsed_hours": 5.270239432255427, "session_id": 2, "event": "train_step", "step": 117450, "epoch": 1, "loss": 1.7299065065383912, "loss_raw": 1.6394530534744263, "perplexity": 5.640126568820471, "muon_lr": 0.002032264232371808, "adamw_lr": 6.096792697115423e-05, "tokens_per_sec": 814777.3336309866, "vram_gb": 2.785593032836914, "grad_norm": 0.095703125, "tokens_processed": 15394406400, "tokens_processed_billions": 15.3944064}
+{"timestamp": 1787646237.2216053, "elapsed_hours": 5.272309639586343, "session_id": 2, "event": "train_step", "step": 117460, "epoch": 1, "loss": 1.7235342156887055, "loss_raw": 1.021685242652893, "perplexity": 5.604300310793285, "muon_lr": 0.0020317772688232586, "adamw_lr": 6.0953318064697746e-05, "tokens_per_sec": 814525.4168875543, "vram_gb": 2.785593032836914, "grad_norm": 0.09765625, "tokens_processed": 15395717120, "tokens_processed_billions": 15.39571712}
+{"timestamp": 1787646244.6739504, "elapsed_hours": 5.274379735456573, "session_id": 2, "event": "train_step", "step": 117470, "epoch": 1, "loss": 1.703272249698639, "loss_raw": 1.2186509370803833, "perplexity": 5.491888852891399, "muon_lr": 0.00203129040528928, "adamw_lr": 6.093871215867839e-05, "tokens_per_sec": 814273.7135087459, "vram_gb": 2.785593032836914, "grad_norm": 0.0966796875, "tokens_processed": 15397027840, "tokens_processed_billions": 15.39702784}
+{"timestamp": 1787646252.23248, "elapsed_hours": 5.276479327016407, "session_id": 2, "event": "train_step", "step": 117480, "epoch": 1, "loss": 1.7108832311630249, "loss_raw": 2.1299185752868652, "perplexity": 5.533846985899186, "muon_lr": 0.0020303169783211844, "adamw_lr": 6.090950934963552e-05, "tokens_per_sec": 814017.6401930492, "vram_gb": 2.363645076751709, "grad_norm": 0.09814453125, "tokens_processed": 15398338560, "tokens_processed_billions": 15.39833856}
+{"timestamp": 1787646259.6817775, "elapsed_hours": 5.278548576368226, "session_id": 2, "event": "train_step", "step": 117490, "epoch": 1, "loss": 1.7209607863426208, "loss_raw": 1.6488502025604248, "perplexity": 5.589896581348345, "muon_lr": 0.00202983041491514, "adamw_lr": 6.0894912447454185e-05, "tokens_per_sec": 813766.4712170495, "vram_gb": 2.785593032836914, "grad_norm": 0.099609375, "tokens_processed": 15399649280, "tokens_processed_billions": 15.39964928}
+{"timestamp": 1787646267.140361, "elapsed_hours": 5.280620405144162, "session_id": 2, "event": "train_step", "step": 117500, "epoch": 1, "loss": 1.6999565887451171, "loss_raw": 1.7639094591140747, "perplexity": 5.473709765959616, "muon_lr": 0.0020293439515798057, "adamw_lr": 6.0880318547394164e-05, "tokens_per_sec": 813515.1031664421, "vram_gb": 2.785593032836914, "grad_norm": 0.09521484375, "tokens_processed": 15400960000, "tokens_processed_billions": 15.40096}
+{"timestamp": 1787646267.1421251, "elapsed_hours": 5.280620895160569, "session_id": 2, "event": "lambda_stats", "step": 117500, "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": 1787646292.8100731, "elapsed_hours": 5.287750880718232, "session_id": 2, "event": "validation", "step": 117500, "epoch": 1, "val_loss": 1.7533713692426682, "val_perplexity": 5.774036307037264, "is_best": false, "best_loss": 1.6233880519866943, "best_perplexity": 5.070239483980759, "train_loss": 1.7639094591140747, "train_loss_avg": 1.6999565887451171}