Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -2597,3 +2597,31 @@
 {"timestamp": 1787645866.6632266, "elapsed_hours": 5.169376756681336, "session_id": 2, "event": "train_step", "step": 117000, "epoch": 1, "loss": 1.7966169464588164, "loss_raw": 1.7359117269515991, "perplexity": 6.029215763391616, "muon_lr": 0.002059693379817389, "adamw_lr": 6.179080139452166e-05, "tokens_per_sec": 827559.6504293596, "vram_gb": 2.363645076751709, "grad_norm": 0.0986328125, "tokens_processed": 15335424000, "tokens_processed_billions": 15.335424}
 {"timestamp": 1787645866.6648788, "elapsed_hours": 5.169377215637101, "session_id": 2, "event": "lambda_stats", "step": 117000, "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": 1787645892.309524, "elapsed_hours": 5.176500728196568, "session_id": 2, "event": "validation", "step": 117000, "epoch": 1, "val_loss": 1.7535317534208297, "val_perplexity": 5.774962445372025, "is_best": false, "best_loss": 1.6233880519866943, "best_perplexity": 5.070239483980759, "train_loss": 1.7359117269515991, "train_loss_avg": 1.7966169464588164}
+{"timestamp": 1787645900.3503673, "elapsed_hours": 5.178734295699331, "session_id": 2, "event": "train_step", "step": 117010, "epoch": 1, "loss": 1.8032986545562744, "loss_raw": 2.228689670562744, "perplexity": 6.0696361112809125, "muon_lr": 0.0020592008379494526, "adamw_lr": 6.177602513848357e-05, "tokens_per_sec": 826128.5719238836, "vram_gb": 2.785593032836914, "grad_norm": 0.0927734375, "tokens_processed": 15336734720, "tokens_processed_billions": 15.33673472}
+{"timestamp": 1787645907.8128836, "elapsed_hours": 5.180807217028406, "session_id": 2, "event": "train_step", "step": 117020, "epoch": 1, "loss": 1.8337109875679016, "loss_raw": 2.46413516998291, "perplexity": 6.257063512324174, "muon_lr": 0.00205870839530526, "adamw_lr": 6.17612518591578e-05, "tokens_per_sec": 825867.1973075905, "vram_gb": 2.785593032836914, "grad_norm": 0.0966796875, "tokens_processed": 15338045440, "tokens_processed_billions": 15.33804544}
+{"timestamp": 1787645915.2599006, "elapsed_hours": 5.182875832782852, "session_id": 2, "event": "train_step", "step": 117030, "epoch": 1, "loss": 1.7390569353103638, "loss_raw": 1.9896430969238281, "perplexity": 5.691972991995133, "muon_lr": 0.002058216051899012, "adamw_lr": 6.174648155697034e-05, "tokens_per_sec": 825606.7201294042, "vram_gb": 2.785593032836914, "grad_norm": 0.09423828125, "tokens_processed": 15339356160, "tokens_processed_billions": 15.33935616}
+{"timestamp": 1787645922.7933319, "elapsed_hours": 5.184968452718523, "session_id": 2, "event": "train_step", "step": 117040, "epoch": 1, "loss": 1.7111061489582062, "loss_raw": 1.9423434734344482, "perplexity": 5.535080716373283, "muon_lr": 0.0020572316628571403, "adamw_lr": 6.17169498857142e-05, "tokens_per_sec": 825342.6125911026, "vram_gb": 2.363645076751709, "grad_norm": 0.09423828125, "tokens_processed": 15340666880, "tokens_processed_billions": 15.34066688}
+{"timestamp": 1787645930.2368436, "elapsed_hours": 5.187036094665527, "session_id": 2, "event": "train_step", "step": 117050, "epoch": 1, "loss": 1.70295867562294, "loss_raw": 1.6354526281356812, "perplexity": 5.49016700889744, "muon_lr": 0.002056739617249904, "adamw_lr": 6.17021885174971e-05, "tokens_per_sec": 825082.7134336784, "vram_gb": 2.785593032836914, "grad_norm": 0.09814453125, "tokens_processed": 15341977600, "tokens_processed_billions": 15.3419776}
+{"timestamp": 1787645937.6843202, "elapsed_hours": 5.189104838238822, "session_id": 2, "event": "train_step", "step": 117060, "epoch": 1, "loss": 1.6913150835037232, "loss_raw": 1.7846465110778809, "perplexity": 5.426612463206137, "muon_lr": 0.002056247670937388, "adamw_lr": 6.168743012812162e-05, "tokens_per_sec": 824822.84647485, "vram_gb": 2.785593032836914, "grad_norm": 0.09814453125, "tokens_processed": 15343288320, "tokens_processed_billions": 15.34328832}
+{"timestamp": 1787645945.1329138, "elapsed_hours": 5.191173892087407, "session_id": 2, "event": "train_step", "step": 117070, "epoch": 1, "loss": 1.64992293715477, "loss_raw": 0.786003053188324, "perplexity": 5.2065785779602125, "muon_lr": 0.0020557558239337756, "adamw_lr": 6.167267471801326e-05, "tokens_per_sec": 824563.1326530459, "vram_gb": 2.785593032836914, "grad_norm": 0.091796875, "tokens_processed": 15344599040, "tokens_processed_billions": 15.34459904}
+{"timestamp": 1787645952.688311, "elapsed_hours": 5.193272613419427, "session_id": 2, "event": "train_step", "step": 117080, "epoch": 1, "loss": 1.6704303681850434, "loss_raw": 1.6540961265563965, "perplexity": 5.31445447721656, "muon_lr": 0.0020547724279100024, "adamw_lr": 6.164317283730007e-05, "tokens_per_sec": 824298.9004752929, "vram_gb": 2.363645076751709, "grad_norm": 0.091796875, "tokens_processed": 15345909760, "tokens_processed_billions": 15.34590976}
+{"timestamp": 1787645960.1284988, "elapsed_hours": 5.195339332289166, "session_id": 2, "event": "train_step", "step": 117090, "epoch": 1, "loss": 1.6867664754390717, "loss_raw": 1.6917043924331665, "perplexity": 5.401984982831709, "muon_lr": 0.0020542808789182003, "adamw_lr": 6.162842636754599e-05, "tokens_per_sec": 824039.9816147197, "vram_gb": 2.785593032836914, "grad_norm": 0.0966796875, "tokens_processed": 15347220480, "tokens_processed_billions": 15.34722048}
+{"timestamp": 1787645967.5767412, "elapsed_hours": 5.197408288518588, "session_id": 2, "event": "train_step", "step": 117100, "epoch": 1, "loss": 1.650269604921341, "loss_raw": 1.5716679096221924, "perplexity": 5.208383843822974, "muon_lr": 0.0020537894292920206, "adamw_lr": 6.161368287876061e-05, "tokens_per_sec": 823780.916836638, "vram_gb": 2.785593032836914, "grad_norm": 0.0966796875, "tokens_processed": 15348531200, "tokens_processed_billions": 15.3485312}
+{"timestamp": 1787645967.5784492, "elapsed_hours": 5.19740876297156, "session_id": 2, "event": "lambda_stats", "step": 117100, "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": 1787645975.0253325, "elapsed_hours": 5.199477341572444, "session_id": 2, "event": "train_step", "step": 117110, "epoch": 1, "loss": 1.5736106252670288, "loss_raw": 1.527897834777832, "perplexity": 4.8240345681463594, "muon_lr": 0.002053298079045635, "adamw_lr": 6.159894237136904e-05, "tokens_per_sec": 823522.0343983008, "vram_gb": 2.785593032836914, "grad_norm": 0.09033203125, "tokens_processed": 15349841920, "tokens_processed_billions": 15.34984192}
+{"timestamp": 1787645982.5586283, "elapsed_hours": 5.201569923758507, "session_id": 2, "event": "train_step", "step": 117120, "epoch": 1, "loss": 1.6254323136806488, "loss_raw": 1.9960267543792725, "perplexity": 5.08061498184018, "muon_lr": 0.0020523156767489254, "adamw_lr": 6.156947030246775e-05, "tokens_per_sec": 823259.6233295295, "vram_gb": 2.363645076751709, "grad_norm": 0.09765625, "tokens_processed": 15351152640, "tokens_processed_billions": 15.35115264}
+{"timestamp": 1787645989.9984417, "elapsed_hours": 5.203636538651255, "session_id": 2, "event": "train_step", "step": 117130, "epoch": 1, "loss": 1.6248618423938752, "loss_raw": 1.3344054222106934, "perplexity": 5.077717463427997, "muon_lr": 0.0020518246247269315, "adamw_lr": 6.155473874180793e-05, "tokens_per_sec": 823001.5509436653, "vram_gb": 2.785593032836914, "grad_norm": 0.09716796875, "tokens_processed": 15352463360, "tokens_processed_billions": 15.35246336}
+{"timestamp": 1787645997.4376485, "elapsed_hours": 5.2057029849953125, "session_id": 2, "event": "train_step", "step": 117140, "epoch": 1, "loss": 1.6674024331569672, "loss_raw": 1.7157483100891113, "perplexity": 5.298386992276191, "muon_lr": 0.002051333672141392, "adamw_lr": 6.154001016424176e-05, "tokens_per_sec": 822743.7092071659, "vram_gb": 2.785593032836914, "grad_norm": 0.09375, "tokens_processed": 15353774080, "tokens_processed_billions": 15.35377408}
+{"timestamp": 1787646004.872814, "elapsed_hours": 5.2077683086527715, "session_id": 2, "event": "train_step", "step": 117150, "epoch": 1, "loss": 1.6619509696960448, "loss_raw": 1.6446926593780518, "perplexity": 5.269581616250979, "muon_lr": 0.0020508428190064653, "adamw_lr": 6.152528457019395e-05, "tokens_per_sec": 822486.2496003599, "vram_gb": 2.785593032836914, "grad_norm": 0.09130859375, "tokens_processed": 15355084800, "tokens_processed_billions": 15.3550848}
+{"timestamp": 1787646012.41024, "elapsed_hours": 5.2098620380957925, "session_id": 2, "event": "train_step", "step": 117160, "epoch": 1, "loss": 1.7104253947734833, "loss_raw": 2.126694440841675, "perplexity": 5.531313969272726, "muon_lr": 0.0020498614111450686, "adamw_lr": 6.149584233435204e-05, "tokens_per_sec": 822224.494288823, "vram_gb": 2.363645076751709, "grad_norm": 0.10986328125, "tokens_processed": 15356395520, "tokens_processed_billions": 15.35639552}
+{"timestamp": 1787646019.85425, "elapsed_hours": 5.211929818789164, "session_id": 2, "event": "train_step", "step": 117170, "epoch": 1, "loss": 1.705987583398819, "loss_raw": 1.859382152557373, "perplexity": 5.50682142805872, "muon_lr": 0.0020493708564468995, "adamw_lr": 6.148112569340698e-05, "tokens_per_sec": 821967.0622926743, "vram_gb": 2.785593032836914, "grad_norm": 0.1484375, "tokens_processed": 15357706240, "tokens_processed_billions": 15.35770624}
+{"timestamp": 1787646027.3192954, "elapsed_hours": 5.214003442459636, "session_id": 2, "event": "train_step", "step": 117180, "epoch": 1, "loss": 1.6946074998378753, "loss_raw": 1.7297675609588623, "perplexity": 5.444508575278804, "muon_lr": 0.0020488804012559446, "adamw_lr": 6.146641203767832e-05, "tokens_per_sec": 821708.9064835313, "vram_gb": 2.785593032836914, "grad_norm": 0.0927734375, "tokens_processed": 15359016960, "tokens_processed_billions": 15.35901696}
+{"timestamp": 1787646034.7644231, "elapsed_hours": 5.216071533428298, "session_id": 2, "event": "train_step", "step": 117190, "epoch": 1, "loss": 1.6555646431446076, "loss_raw": 1.0758253335952759, "perplexity": 5.236035579248534, "muon_lr": 0.00204839004558635, "adamw_lr": 6.145170136759049e-05, "tokens_per_sec": 821451.8384492175, "vram_gb": 2.785593032836914, "grad_norm": 0.09423828125, "tokens_processed": 15360327680, "tokens_processed_billions": 15.36032768}
+{"timestamp": 1787646042.2999454, "elapsed_hours": 5.218164733979437, "session_id": 2, "event": "train_step", "step": 117200, "epoch": 1, "loss": 1.7034268963336945, "loss_raw": 2.122009038925171, "perplexity": 5.4927382206968405, "muon_lr": 0.0020474096328677955, "adamw_lr": 6.142228898603387e-05, "tokens_per_sec": 821190.9956150759, "vram_gb": 2.363645076751709, "grad_norm": 0.095703125, "tokens_processed": 15361638400, "tokens_processed_billions": 15.3616384}
+{"timestamp": 1787646042.3016553, "elapsed_hours": 5.218165209094684, "session_id": 2, "event": "lambda_stats", "step": 117200, "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": 1787646049.7448294, "elapsed_hours": 5.220232757528623, "session_id": 2, "event": "train_step", "step": 117210, "epoch": 1, "loss": 1.7496249878406525, "loss_raw": 1.8240543603897095, "perplexity": 5.752445034626177, "muon_lr": 0.00204691957584711, "adamw_lr": 6.14075872754133e-05, "tokens_per_sec": 820934.3517170174, "vram_gb": 2.785593032836914, "grad_norm": 0.09619140625, "tokens_processed": 15362949120, "tokens_processed_billions": 15.36294912}
+{"timestamp": 1787646057.1870182, "elapsed_hours": 5.222300032046106, "session_id": 2, "event": "train_step", "step": 117220, "epoch": 1, "loss": 1.748749223947525, "loss_raw": 2.5738632678985596, "perplexity": 5.747409456278342, "muon_lr": 0.0020464296184043284, "adamw_lr": 6.139288855212985e-05, "tokens_per_sec": 820678.0188708396, "vram_gb": 2.785593032836914, "grad_norm": 0.09521484375, "tokens_processed": 15364259840, "tokens_processed_billions": 15.36425984}
+{"timestamp": 1787646064.6322424, "elapsed_hours": 5.224368149903086, "session_id": 2, "event": "train_step", "step": 117230, "epoch": 1, "loss": 1.7251575887203217, "loss_raw": 1.6316875219345093, "perplexity": 5.613405569394914, "muon_lr": 0.0020459397605535785, "adamw_lr": 6.137819281660735e-05, "tokens_per_sec": 820421.760488157, "vram_gb": 2.785593032836914, "grad_norm": 0.0947265625, "tokens_processed": 15365570560, "tokens_processed_billions": 15.36557056}
+{"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}