Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -5139,3 +5139,59 @@
 {"timestamp": 1787546569.5658875, "elapsed_hours": 5.540465938912497, "session_id": 2, "event": "train_step", "step": 45500, "epoch": 1, "loss": 1.911975017786026, "loss_raw": 2.7600529193878174, "perplexity": 6.7664394485649, "muon_lr": 0.008214478270405081, "adamw_lr": 0.00024643434811215235, "tokens_per_sec": 149584.74540529266, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 2981888000, "tokens_processed_billions": 2.981888}
 {"timestamp": 1787546569.5711746, "elapsed_hours": 5.540467407968309, "session_id": 2, "event": "lambda_stats", "step": 45500, "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": 1787546583.022972, "elapsed_hours": 5.544204018447134, "session_id": 2, "event": "validation", "step": 45500, "epoch": 1, "val_loss": 1.9575361496210097, "val_perplexity": 7.081856916230893, "is_best": false, "best_loss": 1.9518783861398696, "best_perplexity": 7.041902577470059, "train_loss": 2.7600529193878174, "train_loss_avg": 1.911975017786026}
+{"timestamp": 1787546587.7859836, "elapsed_hours": 5.545527076721191, "session_id": 2, "event": "train_step", "step": 45510, "epoch": 1, "loss": 1.8953964293003083, "loss_raw": 2.173069477081299, "perplexity": 6.655186192555853, "muon_lr": 0.008213868746047882, "adamw_lr": 0.0002464160623814364, "tokens_per_sec": 149480.99567811977, "vram_gb": 27.710740089416504, "grad_norm": 0.19140625, "tokens_processed": 2982543360, "tokens_processed_billions": 2.98254336}
+{"timestamp": 1787546591.909436, "elapsed_hours": 5.546672480172581, "session_id": 2, "event": "train_step", "step": 45520, "epoch": 1, "loss": 1.9893514573574067, "loss_raw": 2.633798360824585, "perplexity": 7.310790864863241, "muon_lr": 0.008212649462568205, "adamw_lr": 0.0002463794838770461, "tokens_per_sec": 149482.94797452385, "vram_gb": 27.299458026885986, "grad_norm": 0.1689453125, "tokens_processed": 2983198720, "tokens_processed_billions": 2.98319872}
+{"timestamp": 1787546595.8836563, "elapsed_hours": 5.547776430249214, "session_id": 2, "event": "train_step", "step": 45530, "epoch": 1, "loss": 1.9786526370048523, "loss_raw": 1.5011001825332642, "perplexity": 7.232990952518705, "muon_lr": 0.008212039703480887, "adamw_lr": 0.00024636119110442656, "tokens_per_sec": 149486.01886354518, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 2983854080, "tokens_processed_billions": 2.98385408}
+{"timestamp": 1787546599.8552191, "elapsed_hours": 5.548879642221663, "session_id": 2, "event": "train_step", "step": 45540, "epoch": 1, "loss": 2.011571868658066, "loss_raw": 1.42544686794281, "perplexity": 7.475057926849784, "muon_lr": 0.008211429866185366, "adamw_lr": 0.0002463428959855609, "tokens_per_sec": 149489.10792974065, "vram_gb": 27.710740089416504, "grad_norm": 0.1328125, "tokens_processed": 2984509440, "tokens_processed_billions": 2.98450944}
+{"timestamp": 1787546603.8057728, "elapsed_hours": 5.549977018303341, "session_id": 2, "event": "train_step", "step": 45550, "epoch": 1, "loss": 2.0043932020664217, "loss_raw": 2.059577465057373, "perplexity": 7.42158912520797, "muon_lr": 0.00821081995069922, "adamw_lr": 0.00024632459852097656, "tokens_per_sec": 149492.35261003574, "vram_gb": 27.710740089416504, "grad_norm": 0.1611328125, "tokens_processed": 2985164800, "tokens_processed_billions": 2.9851648}
+{"timestamp": 1787546607.9466038, "elapsed_hours": 5.551127249002457, "session_id": 2, "event": "train_step", "step": 45560, "epoch": 1, "loss": 2.008723056316376, "loss_raw": 1.4260059595108032, "perplexity": 7.453793193571799, "muon_lr": 0.008209599885225422, "adamw_lr": 0.00024628799655676266, "tokens_per_sec": 149494.1712064711, "vram_gb": 27.299458026885986, "grad_norm": 0.1455078125, "tokens_processed": 2985820160, "tokens_processed_billions": 2.98582016}
+{"timestamp": 1787546612.0018413, "elapsed_hours": 5.55225370393859, "session_id": 2, "event": "train_step", "step": 45570, "epoch": 1, "loss": 1.9773458385467528, "loss_raw": 2.066584825515747, "perplexity": 7.223545064374783, "muon_lr": 0.008208989735272954, "adamw_lr": 0.00024626969205818856, "tokens_per_sec": 149496.63075576304, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 2986475520, "tokens_processed_billions": 2.98647552}
+{"timestamp": 1787546615.98821, "elapsed_hours": 5.553361028499073, "session_id": 2, "event": "train_step", "step": 45580, "epoch": 1, "loss": 2.0549340295791625, "loss_raw": 2.8738770484924316, "perplexity": 7.806322869243713, "muon_lr": 0.008208379507200231, "adamw_lr": 0.0002462513852160069, "tokens_per_sec": 149499.6043810131, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 2987130880, "tokens_processed_billions": 2.98713088}
+{"timestamp": 1787546619.9550626, "elapsed_hours": 5.554462932149569, "session_id": 2, "event": "train_step", "step": 45590, "epoch": 1, "loss": 2.034394301176071, "loss_raw": 1.8838181495666504, "perplexity": 7.647618568777411, "muon_lr": 0.008207769201024853, "adamw_lr": 0.0002462330760307455, "tokens_per_sec": 149502.72318268943, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 2987786240, "tokens_processed_billions": 2.98778624}
+{"timestamp": 1787546624.0662274, "elapsed_hours": 5.555604922241635, "session_id": 2, "event": "train_step", "step": 45600, "epoch": 1, "loss": 1.9724550998210908, "loss_raw": 1.722083568572998, "perplexity": 7.188302843287091, "muon_lr": 0.008206548354436525, "adamw_lr": 0.0002461964506330957, "tokens_per_sec": 149504.759756785, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 2988441600, "tokens_processed_billions": 2.9884416}
+{"timestamp": 1787546624.0692384, "elapsed_hours": 5.555605759090848, "session_id": 2, "event": "lambda_stats", "step": 45600, "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": 1787546628.0395935, "elapsed_hours": 5.556708635158009, "session_id": 2, "event": "train_step", "step": 45610, "epoch": 1, "loss": 2.0059996509552, "loss_raw": 1.3325154781341553, "perplexity": 7.433521110307181, "muon_lr": 0.008205937814058782, "adamw_lr": 0.00024617813442176343, "tokens_per_sec": 149507.82771447304, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 2989096960, "tokens_processed_billions": 2.98909696}
+{"timestamp": 1787546632.0583081, "elapsed_hours": 5.557824944655101, "session_id": 2, "event": "train_step", "step": 45620, "epoch": 1, "loss": 1.9305379569530488, "loss_raw": 1.7143490314483643, "perplexity": 6.893217498601792, "muon_lr": 0.008205327195648796, "adamw_lr": 0.0002461598158694638, "tokens_per_sec": 149510.55429173342, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 2989752320, "tokens_processed_billions": 2.98975232}
+{"timestamp": 1787546636.0078666, "elapsed_hours": 5.5589220442374545, "session_id": 2, "event": "train_step", "step": 45630, "epoch": 1, "loss": 1.9124986732006073, "loss_raw": 2.569552421569824, "perplexity": 6.7699836591111096, "muon_lr": 0.008204716499224173, "adamw_lr": 0.00024614149497672515, "tokens_per_sec": 149513.79683861783, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 2990407680, "tokens_processed_billions": 2.99040768}
+{"timestamp": 1787546640.1399574, "elapsed_hours": 5.560069847371843, "session_id": 2, "event": "train_step", "step": 45640, "epoch": 1, "loss": 1.989376665353775, "loss_raw": 1.3286786079406738, "perplexity": 7.310975157575627, "muon_lr": 0.008203494872401466, "adamw_lr": 0.0002461048461720439, "tokens_per_sec": 149515.67402473613, "vram_gb": 27.299458026885986, "grad_norm": 0.1533203125, "tokens_processed": 2991063040, "tokens_processed_billions": 2.99106304}
+{"timestamp": 1787546644.0986872, "elapsed_hours": 5.561169494390487, "session_id": 2, "event": "train_step", "step": 45650, "epoch": 1, "loss": 2.0329647743701935, "loss_raw": 3.0335805416107178, "perplexity": 7.63669390344361, "muon_lr": 0.00820288394203861, "adamw_lr": 0.00024608651826115823, "tokens_per_sec": 149518.84649156377, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 2991718400, "tokens_processed_billions": 2.9917184}
+{"timestamp": 1787546648.062346, "elapsed_hours": 5.5622705107265045, "session_id": 2, "event": "train_step", "step": 45660, "epoch": 1, "loss": 1.9496388471126556, "loss_raw": 0.8403022289276123, "perplexity": 7.026149608100954, "muon_lr": 0.008202272933731575, "adamw_lr": 0.0002460681880119472, "tokens_per_sec": 149521.98010565795, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 2992373760, "tokens_processed_billions": 2.99237376}
+{"timestamp": 1787546652.0178285, "elapsed_hours": 5.563369255926874, "session_id": 2, "event": "train_step", "step": 45670, "epoch": 1, "loss": 1.9998774480819703, "loss_raw": 1.3615586757659912, "perplexity": 7.3881506114190465, "muon_lr": 0.00820166184749798, "adamw_lr": 0.00024604985542493935, "tokens_per_sec": 149525.1742446949, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 2993029120, "tokens_processed_billions": 2.99302912}
+{"timestamp": 1787546656.2354572, "elapsed_hours": 5.564540819459491, "session_id": 2, "event": "train_step", "step": 45680, "epoch": 1, "loss": 1.9899551820755006, "loss_raw": 1.4620407819747925, "perplexity": 7.315205902616785, "muon_lr": 0.008200439441321608, "adamw_lr": 0.0002460131832396482, "tokens_per_sec": 149526.40822073922, "vram_gb": 27.299458026885986, "grad_norm": 0.15625, "tokens_processed": 2993684480, "tokens_processed_billions": 2.99368448}
+{"timestamp": 1787546660.189785, "elapsed_hours": 5.565639243986872, "session_id": 2, "event": "train_step", "step": 45690, "epoch": 1, "loss": 1.9154680967330933, "loss_raw": 2.013427972793579, "perplexity": 6.79011648455245, "muon_lr": 0.00819982812141408, "adamw_lr": 0.00024599484364242234, "tokens_per_sec": 149529.6090206376, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 2994339840, "tokens_processed_billions": 2.99433984}
+{"timestamp": 1787546664.165104, "elapsed_hours": 5.566743499040603, "session_id": 2, "event": "train_step", "step": 45700, "epoch": 1, "loss": 1.8424205160140992, "loss_raw": 1.592865228652954, "perplexity": 6.311797592999324, "muon_lr": 0.008199216723650494, "adamw_lr": 0.00024597650170951477, "tokens_per_sec": 149532.6512024087, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 2994995200, "tokens_processed_billions": 2.9949952}
+{"timestamp": 1787546664.1684365, "elapsed_hours": 5.566744424965647, "session_id": 2, "event": "lambda_stats", "step": 45700, "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": 1787546668.127547, "elapsed_hours": 5.567844177749421, "session_id": 2, "event": "train_step", "step": 45710, "epoch": 1, "loss": 1.8918952751159668, "loss_raw": 2.0790910720825195, "perplexity": 6.63192610191239, "muon_lr": 0.008198605248048484, "adamw_lr": 0.00024595815744145446, "tokens_per_sec": 149535.78828243108, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 2995650560, "tokens_processed_billions": 2.99565056}
+{"timestamp": 1787546672.3478074, "elapsed_hours": 5.569016472299894, "session_id": 2, "event": "train_step", "step": 45720, "epoch": 1, "loss": 1.8575853180885316, "loss_raw": 1.605305790901184, "perplexity": 6.408244202783459, "muon_lr": 0.008197382063399719, "adamw_lr": 0.00024592146190199156, "tokens_per_sec": 149536.99933167303, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 2996305920, "tokens_processed_billions": 2.99630592}
+{"timestamp": 1787546676.3766396, "elapsed_hours": 5.570135592420896, "session_id": 2, "event": "train_step", "step": 45730, "epoch": 1, "loss": 1.857103168964386, "loss_raw": 1.9178823232650757, "perplexity": 6.4051552181892255, "muon_lr": 0.008196770354388244, "adamw_lr": 0.00024590311063164724, "tokens_per_sec": 149539.63904530834, "vram_gb": 27.710740089416504, "grad_norm": 0.15234375, "tokens_processed": 2996961280, "tokens_processed_billions": 2.99696128}
+{"timestamp": 1787546680.353416, "elapsed_hours": 5.571240252322621, "session_id": 2, "event": "train_step", "step": 45740, "epoch": 1, "loss": 1.9894866299629212, "loss_raw": 2.754265785217285, "perplexity": 7.3117791503058704, "muon_lr": 0.008196158567608886, "adamw_lr": 0.0002458847570282665, "tokens_per_sec": 149542.6666820624, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2997616640, "tokens_processed_billions": 2.99761664}
+{"timestamp": 1787546684.3552244, "elapsed_hours": 5.572351865768432, "session_id": 2, "event": "train_step", "step": 45750, "epoch": 1, "loss": 2.055776762962341, "loss_raw": 1.9602280855178833, "perplexity": 7.812904290927866, "muon_lr": 0.00819554670307929, "adamw_lr": 0.0002458664010923787, "tokens_per_sec": 149545.50470554363, "vram_gb": 27.710740089416504, "grad_norm": 0.16796875, "tokens_processed": 2998272000, "tokens_processed_billions": 2.998272}
+{"timestamp": 1787546688.4641862, "elapsed_hours": 5.57349324411816, "session_id": 2, "event": "train_step", "step": 45760, "epoch": 1, "loss": 2.0699242889881133, "loss_raw": 2.4858558177948, "perplexity": 7.924223144184805, "muon_lr": 0.008194322740839975, "adamw_lr": 0.0002458296822251992, "tokens_per_sec": 149547.54355629507, "vram_gb": 27.299458026885986, "grad_norm": 0.1611328125, "tokens_processed": 2998927360, "tokens_processed_billions": 2.99892736}
+{"timestamp": 1787546692.4364264, "elapsed_hours": 5.574596644176377, "session_id": 2, "event": "train_step", "step": 45770, "epoch": 1, "loss": 2.1113198935985564, "loss_raw": 1.8256053924560547, "perplexity": 8.259135273369038, "muon_lr": 0.008193710643165548, "adamw_lr": 0.0002458113192949664, "tokens_per_sec": 149550.60114356177, "vram_gb": 27.710740089416504, "grad_norm": 0.251953125, "tokens_processed": 2999582720, "tokens_processed_billions": 2.99958272}
+{"timestamp": 1787546696.3980668, "elapsed_hours": 5.575697099765142, "session_id": 2, "event": "train_step", "step": 45780, "epoch": 1, "loss": 2.1326808202266694, "loss_raw": 2.3513612747192383, "perplexity": 8.437455821602088, "muon_lr": 0.008193098467811476, "adamw_lr": 0.00024579295403434425, "tokens_per_sec": 149553.7368299605, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 3000238080, "tokens_processed_billions": 3.00023808}
+{"timestamp": 1787546700.371244, "elapsed_hours": 5.576800760295656, "session_id": 2, "event": "train_step", "step": 45790, "epoch": 1, "loss": 2.0192677330970765, "loss_raw": 2.0197231769561768, "perplexity": 7.532806888397048, "muon_lr": 0.008192486214795415, "adamw_lr": 0.00024577458644386237, "tokens_per_sec": 149556.78399966727, "vram_gb": 27.710740089416504, "grad_norm": 0.166015625, "tokens_processed": 3000893440, "tokens_processed_billions": 3.00089344}
+{"timestamp": 1787546704.496138, "elapsed_hours": 5.577946564091577, "session_id": 2, "event": "train_step", "step": 45800, "epoch": 1, "loss": 2.00995055437088, "loss_raw": 1.259878158569336, "perplexity": 7.462948328020906, "muon_lr": 0.008191261475847943, "adamw_lr": 0.00024573784427543824, "tokens_per_sec": 149558.7002844038, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 3001548800, "tokens_processed_billions": 3.0015488}
+{"timestamp": 1787546704.499525, "elapsed_hours": 5.577947505116462, "session_id": 2, "event": "lambda_stats", "step": 45800, "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": 1787546708.5248325, "elapsed_hours": 5.579065645933151, "session_id": 2, "event": "train_step", "step": 45810, "epoch": 1, "loss": 1.9715379965305329, "loss_raw": 2.1239776611328125, "perplexity": 7.181713449135365, "muon_lr": 0.008190648989951851, "adamw_lr": 0.0002457194696985555, "tokens_per_sec": 149561.33256074344, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 3002204160, "tokens_processed_billions": 3.00220416}
+{"timestamp": 1787546712.4978151, "elapsed_hours": 5.580169252223438, "session_id": 2, "event": "train_step", "step": 45820, "epoch": 1, "loss": 1.8772112548351287, "loss_raw": 1.4204027652740479, "perplexity": 6.535254267151146, "muon_lr": 0.008190036426464407, "adamw_lr": 0.0002457010927939321, "tokens_per_sec": 149564.37871875655, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 3002859520, "tokens_processed_billions": 3.00285952}
+{"timestamp": 1787546716.5199516, "elapsed_hours": 5.581286512348387, "session_id": 2, "event": "train_step", "step": 45830, "epoch": 1, "loss": 1.826011688709259, "loss_raw": 0.9916247725486755, "perplexity": 6.209073492186197, "muon_lr": 0.008189423785403271, "adamw_lr": 0.0002456827135620981, "tokens_per_sec": 149567.05743776998, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 3003514880, "tokens_processed_billions": 3.00351488}
+{"timestamp": 1787546720.7242289, "elapsed_hours": 5.582454367147552, "session_id": 2, "event": "train_step", "step": 45840, "epoch": 1, "loss": 1.8311389935016633, "loss_raw": 1.6773253679275513, "perplexity": 6.240991060084439, "muon_lr": 0.0081881982706306, "adamw_lr": 0.00024564594811891797, "tokens_per_sec": 149568.3785895533, "vram_gb": 27.299458026885986, "grad_norm": 0.150390625, "tokens_processed": 3004170240, "tokens_processed_billions": 3.00417024}
+{"timestamp": 1787546724.6953595, "elapsed_hours": 5.5835574589835275, "session_id": 2, "event": "train_step", "step": 45850, "epoch": 1, "loss": 1.8739649498462676, "loss_raw": 3.352648973464966, "perplexity": 6.514073237263061, "muon_lr": 0.008187585396954408, "adamw_lr": 0.00024562756190863214, "tokens_per_sec": 149571.43510856765, "vram_gb": 27.710740089416504, "grad_norm": 0.158203125, "tokens_processed": 3004825600, "tokens_processed_billions": 3.0048256}
+{"timestamp": 1787546728.6513202, "elapsed_hours": 5.5846563370360265, "session_id": 2, "event": "train_step", "step": 45860, "epoch": 1, "loss": 1.9057636415958406, "loss_raw": 2.338393449783325, "perplexity": 6.7245408065038195, "muon_lr": 0.008186972445775204, "adamw_lr": 0.00024560917337325606, "tokens_per_sec": 149574.60370109338, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 3005480960, "tokens_processed_billions": 3.00548096}
+{"timestamp": 1787546732.6153324, "elapsed_hours": 5.585757451388571, "session_id": 2, "event": "train_step", "step": 45870, "epoch": 1, "loss": 1.9289757931232452, "loss_raw": 2.971782684326172, "perplexity": 6.8824575701282145, "muon_lr": 0.008186359417110671, "adamw_lr": 0.00024559078251332005, "tokens_per_sec": 149577.71033852425, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 3006136320, "tokens_processed_billions": 3.00613632}
+{"timestamp": 1787546736.8268807, "elapsed_hours": 5.58692732612292, "session_id": 2, "event": "train_step", "step": 45880, "epoch": 1, "loss": 1.9147981238365173, "loss_raw": 0.9926398992538452, "perplexity": 6.785568814121278, "muon_lr": 0.00818513312739632, "adamw_lr": 0.00024555399382188956, "tokens_per_sec": 149578.97396839407, "vram_gb": 27.299458026885986, "grad_norm": 0.1455078125, "tokens_processed": 3006791680, "tokens_processed_billions": 3.00679168}
+{"timestamp": 1787546740.8023078, "elapsed_hours": 5.588031611442566, "session_id": 2, "event": "train_step", "step": 45890, "epoch": 1, "loss": 1.9169181096553802, "loss_raw": 1.7449837923049927, "perplexity": 6.799969382887498, "muon_lr": 0.008184519866381865, "adamw_lr": 0.0002455355959914559, "tokens_per_sec": 149581.9940043964, "vram_gb": 27.710740089416504, "grad_norm": 0.1474609375, "tokens_processed": 3007447040, "tokens_processed_billions": 3.00744704}
+{"timestamp": 1787546744.757692, "elapsed_hours": 5.5891303290923435, "session_id": 2, "event": "train_step", "step": 45900, "epoch": 1, "loss": 1.8901990127563477, "loss_raw": 0.9771864414215088, "perplexity": 6.620686150941883, "muon_lr": 0.008183906527952806, "adamw_lr": 0.0002455171958385841, "tokens_per_sec": 149585.1619012648, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 3008102400, "tokens_processed_billions": 3.0081024}
+{"timestamp": 1787546744.7609234, "elapsed_hours": 5.589131226870749, "session_id": 2, "event": "lambda_stats", "step": 45900, "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": 1787546748.7431276, "elapsed_hours": 5.590237394505077, "session_id": 2, "event": "train_step", "step": 45910, "epoch": 1, "loss": 1.8920174479484557, "loss_raw": 1.879597783088684, "perplexity": 6.632736392605865, "muon_lr": 0.008183293112126826, "adamw_lr": 0.00024549879336380475, "tokens_per_sec": 149588.10563242596, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 3008757760, "tokens_processed_billions": 3.00875776}
+{"timestamp": 1787546752.882965, "elapsed_hours": 5.5913873494333695, "session_id": 2, "event": "train_step", "step": 45920, "epoch": 1, "loss": 1.9230578529834748, "loss_raw": 1.2063660621643066, "perplexity": 6.841847879591895, "muon_lr": 0.008182066048354869, "adamw_lr": 0.000245461981450646, "tokens_per_sec": 149589.89857620953, "vram_gb": 27.299458026885986, "grad_norm": 0.138671875, "tokens_processed": 3009413120, "tokens_processed_billions": 3.00941312}
+{"timestamp": 1787546756.942133, "elapsed_hours": 5.592514895995458, "session_id": 2, "event": "train_step", "step": 45930, "epoch": 1, "loss": 1.9326160216331483, "loss_raw": 0.9489518404006958, "perplexity": 6.907556944405576, "muon_lr": 0.008181452400444277, "adamw_lr": 0.0002454435720133283, "tokens_per_sec": 149592.2921107122, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 3010068480, "tokens_processed_billions": 3.01006848}
+{"timestamp": 1787546760.9881756, "elapsed_hours": 5.593638796740108, "session_id": 2, "event": "train_step", "step": 45940, "epoch": 1, "loss": 1.915353296995163, "loss_raw": 2.402801752090454, "perplexity": 6.789337025701199, "muon_lr": 0.008180838675207535, "adamw_lr": 0.000245425160256226, "tokens_per_sec": 149594.7816928342, "vram_gb": 27.710740089416504, "grad_norm": 0.13671875, "tokens_processed": 3010723840, "tokens_processed_billions": 3.01072384}
+{"timestamp": 1787546765.0225701, "elapsed_hours": 5.594759461945958, "session_id": 2, "event": "train_step", "step": 45950, "epoch": 1, "loss": 1.9337988233566283, "loss_raw": 2.633063554763794, "perplexity": 6.915732048474953, "muon_lr": 0.008180224872662345, "adamw_lr": 0.0002454067461798703, "tokens_per_sec": 149597.35621858685, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 3011379200, "tokens_processed_billions": 3.0113792}
+{"timestamp": 1787546769.1553032, "elapsed_hours": 5.595907443298234, "session_id": 2, "event": "train_step", "step": 45960, "epoch": 1, "loss": 2.001968412399292, "loss_raw": 2.8266425132751465, "perplexity": 7.403615132965479, "muon_lr": 0.008178997035717415, "adamw_lr": 0.0002453699110715224, "tokens_per_sec": 149599.19928296507, "vram_gb": 27.299458026885986, "grad_norm": 0.154296875, "tokens_processed": 3012034560, "tokens_processed_billions": 3.01203456}
+{"timestamp": 1787546773.2226708, "elapsed_hours": 5.597037267552482, "session_id": 2, "event": "train_step", "step": 45970, "epoch": 1, "loss": 1.9905427014827728, "loss_raw": 0.6366581916809082, "perplexity": 7.31950499082798, "muon_lr": 0.008178383001353085, "adamw_lr": 0.0002453514900405925, "tokens_per_sec": 149601.5282774068, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 3012689920, "tokens_processed_billions": 3.01268992}
+{"timestamp": 1787546777.2110574, "elapsed_hours": 5.5981451528602175, "session_id": 2, "event": "train_step", "step": 45980, "epoch": 1, "loss": 2.0737397050857544, "loss_raw": 2.553571939468384, "perplexity": 7.954515104198689, "muon_lr": 0.008177768889751122, "adamw_lr": 0.0002453330666925336, "tokens_per_sec": 149604.44218066736, "vram_gb": 27.710740089416504, "grad_norm": 2.109375, "tokens_processed": 3013345280, "tokens_processed_billions": 3.01334528}
+{"timestamp": 1787546781.2414975, "elapsed_hours": 5.599264719552464, "session_id": 2, "event": "train_step", "step": 45990, "epoch": 1, "loss": 2.0895308685302734, "loss_raw": 2.0215044021606445, "perplexity": 8.081123165713898, "muon_lr": 0.00817715470092923, "adamw_lr": 0.00024531464102787683, "tokens_per_sec": 149607.0425974093, "vram_gb": 27.710740089416504, "grad_norm": 0.1396484375, "tokens_processed": 3014000640, "tokens_processed_billions": 3.01400064}
+{"timestamp": 1787546785.393812, "elapsed_hours": 5.6004181402259405, "session_id": 2, "event": "train_step", "step": 46000, "epoch": 1, "loss": 2.0855834531784057, "loss_raw": 1.733067512512207, "perplexity": 8.049286493693309, "muon_lr": 0.008175926091696518, "adamw_lr": 0.0002452777827508955, "tokens_per_sec": 149608.73725842414, "vram_gb": 27.299458026885986, "grad_norm": 0.1572265625, "tokens_processed": 3014656000, "tokens_processed_billions": 3.014656}
+{"timestamp": 1787546785.3969364, "elapsed_hours": 5.600419008268251, "session_id": 2, "event": "lambda_stats", "step": 46000, "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": 1787546798.9118354, "elapsed_hours": 5.604173146883647, "session_id": 2, "event": "validation", "step": 46000, "epoch": 1, "val_loss": 1.9484020507335662, "val_perplexity": 7.017465063320521, "is_best": true, "best_loss": 1.9484020507335662, "best_perplexity": 7.017465063320521, "train_loss": 1.733067512512207, "train_loss_avg": 2.0855834531784057}