Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -3744,3 +3744,42 @@
 {"timestamp": 1787541097.586105, "elapsed_hours": 4.020471554928355, "session_id": 2, "event": "train_step", "step": 33140, "epoch": 1, "loss": 1.9376701939105987, "loss_raw": 2.3959147930145264, "perplexity": 6.942557301630516, "muon_lr": 0.009052475117549957, "adamw_lr": 0.0002715742535264987, "tokens_per_sec": 150172.45644972337, "vram_gb": 27.710740089416504, "grad_norm": 0.1650390625, "tokens_processed": 2171863040, "tokens_processed_billions": 2.17186304}
 {"timestamp": 1787541101.6418967, "elapsed_hours": 4.021598163776927, "session_id": 2, "event": "train_step", "step": 33150, "epoch": 1, "loss": 2.090186823606491, "loss_raw": 2.2243900299072266, "perplexity": 8.086425758417157, "muon_lr": 0.009052005995796693, "adamw_lr": 0.0002715601798739007, "tokens_per_sec": 150175.656407658, "vram_gb": 27.710740089416504, "grad_norm": 0.173828125, "tokens_processed": 2172518400, "tokens_processed_billions": 2.1725184}
 {"timestamp": 1787541105.7880845, "elapsed_hours": 4.022749882472886, "session_id": 2, "event": "train_step", "step": 33160, "epoch": 1, "loss": 2.042172040939331, "loss_raw": 0.9378709197044373, "perplexity": 7.707331671280303, "muon_lr": 0.009051067445011707, "adamw_lr": 0.00027153202335035117, "tokens_per_sec": 150177.91597800463, "vram_gb": 27.299458026885986, "grad_norm": 0.138671875, "tokens_processed": 2173173760, "tokens_processed_billions": 2.17317376}
+{"timestamp": 1787541110.4023461, "elapsed_hours": 4.024031622144911, "session_id": 2, "event": "train_step", "step": 33170, "epoch": 1, "loss": 2.080719155073166, "loss_raw": 1.0695220232009888, "perplexity": 8.010227439112052, "muon_lr": 0.009050598016007053, "adamw_lr": 0.00027151794048021154, "tokens_per_sec": 150175.31911463788, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2173829120, "tokens_processed_billions": 2.17382912}
+{"timestamp": 1787541114.3572905, "elapsed_hours": 4.025130217605167, "session_id": 2, "event": "train_step", "step": 33180, "epoch": 1, "loss": 2.069536882638931, "loss_raw": 2.7436468601226807, "perplexity": 7.921153844397912, "muon_lr": 0.009050128484612341, "adamw_lr": 0.0002715038545383701, "tokens_per_sec": 150179.5612944252, "vram_gb": 27.710740089416504, "grad_norm": 0.140625, "tokens_processed": 2174484480, "tokens_processed_billions": 2.17448448}
+{"timestamp": 1787541118.395802, "elapsed_hours": 4.0262520264254675, "session_id": 2, "event": "train_step", "step": 33190, "epoch": 1, "loss": 1.9438152551651, "loss_raw": 2.020371913909912, "perplexity": 6.985351092079683, "muon_lr": 0.009049658850841108, "adamw_lr": 0.0002714897655252332, "tokens_per_sec": 150182.9343558786, "vram_gb": 27.710740089416504, "grad_norm": 0.14453125, "tokens_processed": 2175139840, "tokens_processed_billions": 2.17513984}
+{"timestamp": 1787541122.5163074, "elapsed_hours": 4.027396611173947, "session_id": 2, "event": "train_step", "step": 33200, "epoch": 1, "loss": 1.8787486600875853, "loss_raw": 2.0083601474761963, "perplexity": 6.545309328759345, "muon_lr": 0.00904871927622326, "adamw_lr": 0.0002714615782866978, "tokens_per_sec": 150185.45521819926, "vram_gb": 27.299458026885986, "grad_norm": 0.1630859375, "tokens_processed": 2175795200, "tokens_processed_billions": 2.1757952}
+{"timestamp": 1787541122.521955, "elapsed_hours": 4.027398180100653, "session_id": 2, "event": "lambda_stats", "step": 33200, "epoch": 1, "lambda_min": 1.109375, "lambda_max": 2.015625, "lambda_mean": 1.1966145833333333, "lambda_std": 0.20269468977080954, "lambda_all": [2.015625, 1.265625, 1.2109375, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.125, 1.140625, 1.109375, 1.140625, 1.109375, 1.125, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787541126.531197, "elapsed_hours": 4.028511858317587, "session_id": 2, "event": "train_step", "step": 33210, "epoch": 1, "loss": 1.9463893020153045, "loss_raw": 1.8754003047943115, "perplexity": 7.003354874404752, "muon_lr": 0.009048249335403744, "adamw_lr": 0.0002714474800621122, "tokens_per_sec": 150189.0705919363, "vram_gb": 27.710740089416504, "grad_norm": 0.1611328125, "tokens_processed": 2176450560, "tokens_processed_billions": 2.17645056}
+{"timestamp": 1787541130.669004, "elapsed_hours": 4.02966124912103, "session_id": 2, "event": "train_step", "step": 33220, "epoch": 1, "loss": 1.8973222422599791, "loss_raw": 1.5341739654541016, "perplexity": 6.668015185529997, "muon_lr": 0.009047779292261894, "adamw_lr": 0.0002714333787678567, "tokens_per_sec": 150191.4096300625, "vram_gb": 27.710740089416504, "grad_norm": 0.1474609375, "tokens_processed": 2177105920, "tokens_processed_billions": 2.17710592}
+{"timestamp": 1787541134.7419662, "elapsed_hours": 4.030792627665732, "session_id": 2, "event": "train_step", "step": 33230, "epoch": 1, "loss": 1.8993233013153077, "loss_raw": 2.676063060760498, "perplexity": 6.681371636765049, "muon_lr": 0.009047309146811271, "adamw_lr": 0.0002714192744043381, "tokens_per_sec": 150194.42000668307, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 2177761280, "tokens_processed_billions": 2.17776128}
+{"timestamp": 1787541138.9129632, "elapsed_hours": 4.031951237652037, "session_id": 2, "event": "train_step", "step": 33240, "epoch": 1, "loss": 1.940239404439926, "loss_raw": 3.2845113277435303, "perplexity": 6.9604171259505145, "muon_lr": 0.009046368549037936, "adamw_lr": 0.000271391056471138, "tokens_per_sec": 150196.41096557525, "vram_gb": 27.299458026885986, "grad_norm": 0.146484375, "tokens_processed": 2178416640, "tokens_processed_billions": 2.17841664}
+{"timestamp": 1787541142.933771, "elapsed_hours": 4.033068128824234, "session_id": 2, "event": "train_step", "step": 33250, "epoch": 1, "loss": 1.9645666205883026, "loss_raw": 1.1492345333099365, "perplexity": 7.131821135872311, "muon_lr": 0.009045898096742347, "adamw_lr": 0.00027137694290227035, "tokens_per_sec": 150199.95836893274, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 2179072000, "tokens_processed_billions": 2.179072}
+{"timestamp": 1787541146.9353356, "elapsed_hours": 4.034179674453205, "session_id": 2, "event": "train_step", "step": 33260, "epoch": 1, "loss": 1.9640223276615143, "loss_raw": 2.04266619682312, "perplexity": 7.127940392299383, "muon_lr": 0.009045427542192234, "adamw_lr": 0.000271362826265767, "tokens_per_sec": 150203.70099838858, "vram_gb": 27.710740089416504, "grad_norm": 0.1357421875, "tokens_processed": 2179727360, "tokens_processed_billions": 2.17972736}
+{"timestamp": 1787541150.9107943, "elapsed_hours": 4.035283968581093, "session_id": 2, "event": "train_step", "step": 33270, "epoch": 1, "loss": 1.9907607078552245, "loss_raw": 1.6677042245864868, "perplexity": 7.321100863508059, "muon_lr": 0.009044956885401165, "adamw_lr": 0.00027134870656203493, "tokens_per_sec": 150207.71315279798, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2180382720, "tokens_processed_billions": 2.18038272}
+{"timestamp": 1787541155.0930343, "elapsed_hours": 4.036445701850785, "session_id": 2, "event": "train_step", "step": 33280, "epoch": 1, "loss": 2.0164204382896425, "loss_raw": 1.4986188411712646, "perplexity": 7.511389272061741, "muon_lr": 0.009044015265150453, "adamw_lr": 0.00027132045795451353, "tokens_per_sec": 150209.5834500681, "vram_gb": 27.299458026885986, "grad_norm": 0.158203125, "tokens_processed": 2181038080, "tokens_processed_billions": 2.18103808}
+{"timestamp": 1787541159.0765293, "elapsed_hours": 4.03755222843753, "session_id": 2, "event": "train_step", "step": 33290, "epoch": 1, "loss": 1.9645201766490936, "loss_raw": 1.321237564086914, "perplexity": 7.1314899136967185, "muon_lr": 0.009043544301717964, "adamw_lr": 0.0002713063290515388, "tokens_per_sec": 150213.50847937432, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 2181693440, "tokens_processed_billions": 2.18169344}
+{"timestamp": 1787541163.1346598, "elapsed_hours": 4.038679486910502, "session_id": 2, "event": "train_step", "step": 33300, "epoch": 1, "loss": 1.942128621339798, "loss_raw": 1.94929039478302, "perplexity": 6.973579292794236, "muon_lr": 0.009043073236098823, "adamw_lr": 0.00027129219708296466, "tokens_per_sec": 150216.65911364328, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 2182348800, "tokens_processed_billions": 2.1823488}
+{"timestamp": 1787541163.1458066, "elapsed_hours": 4.0386825834380256, "session_id": 2, "event": "lambda_stats", "step": 33300, "epoch": 1, "lambda_min": 1.109375, "lambda_max": 2.015625, "lambda_mean": 1.1966145833333333, "lambda_std": 0.20269468977080954, "lambda_all": [2.015625, 1.265625, 1.2109375, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.125, 1.140625, 1.109375, 1.140625, 1.109375, 1.125, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787541167.1356294, "elapsed_hours": 4.039790867368381, "session_id": 2, "event": "train_step", "step": 33310, "epoch": 1, "loss": 1.9788974559307098, "loss_raw": 1.8078882694244385, "perplexity": 7.234761942371509, "muon_lr": 0.009042602068306621, "adamw_lr": 0.0002712780620491986, "tokens_per_sec": 150220.39932798978, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 2183004160, "tokens_processed_billions": 2.18300416}
+{"timestamp": 1787541171.3708086, "elapsed_hours": 4.040967305898667, "session_id": 2, "event": "train_step", "step": 33320, "epoch": 1, "loss": 1.9499175345897675, "loss_raw": 1.5025098323822021, "perplexity": 7.028107980883356, "muon_lr": 0.009041659426257377, "adamw_lr": 0.0002712497827877213, "tokens_per_sec": 150221.71670347857, "vram_gb": 27.299458026885986, "grad_norm": 0.1396484375, "tokens_processed": 2183659520, "tokens_processed_billions": 2.18365952}
+{"timestamp": 1787541175.3693497, "elapsed_hours": 4.042078011830648, "session_id": 2, "event": "train_step", "step": 33330, "epoch": 1, "loss": 1.817759521007538, "loss_raw": 1.193668246269226, "perplexity": 6.158046009701082, "muon_lr": 0.009041187952027518, "adamw_lr": 0.0002712356385608255, "tokens_per_sec": 150225.47839251783, "vram_gb": 27.710740089416504, "grad_norm": 0.96875, "tokens_processed": 2184314880, "tokens_processed_billions": 2.18431488}
+{"timestamp": 1787541179.4584126, "elapsed_hours": 4.0432138627105285, "session_id": 2, "event": "train_step", "step": 33340, "epoch": 1, "loss": 1.9084736120700836, "loss_raw": 3.1608710289001465, "perplexity": 6.742788828174561, "muon_lr": 0.009040716375678964, "adamw_lr": 0.0002712214912703688, "tokens_per_sec": 150228.30234576084, "vram_gb": 27.710740089416504, "grad_norm": 0.1533203125, "tokens_processed": 2184970240, "tokens_processed_billions": 2.18497024}
+{"timestamp": 1787541183.4371622, "elapsed_hours": 4.044319070710076, "session_id": 2, "event": "train_step", "step": 33350, "epoch": 1, "loss": 1.9010231912136077, "loss_raw": 2.330165386199951, "perplexity": 6.692738891720851, "muon_lr": 0.009040244697225308, "adamw_lr": 0.0002712073409167592, "tokens_per_sec": 150232.2640522061, "vram_gb": 27.710740089416504, "grad_norm": 0.1474609375, "tokens_processed": 2185625600, "tokens_processed_billions": 2.1856256}
+{"timestamp": 1787541187.5775197, "elapsed_hours": 4.045469170014064, "session_id": 2, "event": "train_step", "step": 33360, "epoch": 1, "loss": 1.8609915745258332, "loss_raw": 0.976794958114624, "perplexity": 6.430109544189752, "muon_lr": 0.009039301034057118, "adamw_lr": 0.0002711790310217135, "tokens_per_sec": 150234.55532819775, "vram_gb": 27.299458026885986, "grad_norm": 0.146484375, "tokens_processed": 2186280960, "tokens_processed_billions": 2.18628096}
+{"timestamp": 1787541191.5525827, "elapsed_hours": 4.046573354270723, "session_id": 2, "event": "train_step", "step": 33370, "epoch": 1, "loss": 1.910047743320465, "loss_raw": 1.742026925086975, "perplexity": 6.753411221110625, "muon_lr": 0.009038829049369791, "adamw_lr": 0.00027116487148109373, "tokens_per_sec": 150238.55179868196, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 2186936320, "tokens_processed_billions": 2.18693632}
+{"timestamp": 1787541195.549757, "elapsed_hours": 4.04768368045489, "session_id": 2, "event": "train_step", "step": 33380, "epoch": 1, "loss": 2.1011359131336214, "loss_raw": 2.108783006668091, "perplexity": 8.175451242617095, "muon_lr": 0.009038356962631795, "adamw_lr": 0.0002711507088789538, "tokens_per_sec": 150242.31722804363, "vram_gb": 27.710740089416504, "grad_norm": 0.193359375, "tokens_processed": 2187591680, "tokens_processed_billions": 2.18759168}
+{"timestamp": 1787541199.584334, "elapsed_hours": 4.048804396324687, "session_id": 2, "event": "train_step", "step": 33390, "epoch": 1, "loss": 2.0079468905925753, "loss_raw": 0.6590254306793213, "perplexity": 7.448010059408237, "muon_lr": 0.009037884773856739, "adamw_lr": 0.00027113654321570213, "tokens_per_sec": 150245.69519733996, "vram_gb": 27.710740089416504, "grad_norm": 0.15625, "tokens_processed": 2188247040, "tokens_processed_billions": 2.18824704}
+{"timestamp": 1787541203.740381, "elapsed_hours": 4.049958853986528, "session_id": 2, "event": "train_step", "step": 33400, "epoch": 1, "loss": 2.0881562173366546, "loss_raw": 2.4770991802215576, "perplexity": 8.070022071922397, "muon_lr": 0.009036940090249919, "adamw_lr": 0.00027110820270749754, "tokens_per_sec": 150247.8177225075, "vram_gb": 27.299458026885986, "grad_norm": 0.1884765625, "tokens_processed": 2188902400, "tokens_processed_billions": 2.1889024}
+{"timestamp": 1787541203.7471182, "elapsed_hours": 4.049960725638601, "session_id": 2, "event": "lambda_stats", "step": 33400, "epoch": 1, "lambda_min": 1.109375, "lambda_max": 2.015625, "lambda_mean": 1.1966145833333333, "lambda_std": 0.20269468977080954, "lambda_all": [2.015625, 1.265625, 1.2109375, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.125, 1.140625, 1.109375, 1.140625, 1.109375, 1.125, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787541207.7840273, "elapsed_hours": 4.051082089013524, "session_id": 2, "event": "train_step", "step": 33410, "epoch": 1, "loss": 2.040396809577942, "loss_raw": 2.0194995403289795, "perplexity": 7.693661511822776, "muon_lr": 0.0090364675954454, "adamw_lr": 0.00027109402786336194, "tokens_per_sec": 150251.09895170064, "vram_gb": 27.710740089416504, "grad_norm": 0.15234375, "tokens_processed": 2189557760, "tokens_processed_billions": 2.18955776}
+{"timestamp": 1787541211.7585855, "elapsed_hours": 4.0521861328019035, "session_id": 2, "event": "train_step", "step": 33420, "epoch": 1, "loss": 2.016288765668869, "loss_raw": 1.0357191562652588, "perplexity": 7.510400292862809, "muon_lr": 0.009035994998658307, "adamw_lr": 0.0002710798499597491, "tokens_per_sec": 150255.0901016463, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2190213120, "tokens_processed_billions": 2.19021312}
+{"timestamp": 1787541215.7842884, "elapsed_hours": 4.05330438375473, "session_id": 2, "event": "train_step", "step": 33430, "epoch": 1, "loss": 1.9874677240848542, "loss_raw": 3.0710511207580566, "perplexity": 7.297032247713362, "muon_lr": 0.009035522299902263, "adamw_lr": 0.0002710656689970679, "tokens_per_sec": 150258.55206321704, "vram_gb": 27.710740089416504, "grad_norm": 0.1640625, "tokens_processed": 2190868480, "tokens_processed_billions": 2.19086848}
+{"timestamp": 1787541219.914308, "elapsed_hours": 4.054451611306932, "session_id": 2, "event": "train_step", "step": 33440, "epoch": 1, "loss": 2.044324280023575, "loss_raw": 1.72783362865448, "perplexity": 7.7239375552446585, "muon_lr": 0.009034576596537874, "adamw_lr": 0.0002710372978961362, "tokens_per_sec": 150260.93675815675, "vram_gb": 27.299458026885986, "grad_norm": 0.22265625, "tokens_processed": 2191523840, "tokens_processed_billions": 2.19152384}
+{"timestamp": 1787541223.9223828, "elapsed_hours": 4.055564965407053, "session_id": 2, "event": "train_step", "step": 33450, "epoch": 1, "loss": 1.9886451876163482, "loss_raw": 2.552006959915161, "perplexity": 7.305629297436183, "muon_lr": 0.009034103591956798, "adamw_lr": 0.0002710231077587038, "tokens_per_sec": 150264.57761782187, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 2192179200, "tokens_processed_billions": 2.1921792}
+{"timestamp": 1787541227.961873, "elapsed_hours": 4.0566870461569895, "session_id": 2, "event": "train_step", "step": 33460, "epoch": 1, "loss": 2.0917638087272645, "loss_raw": 2.5710554122924805, "perplexity": 8.099187991799647, "muon_lr": 0.009033630485461318, "adamw_lr": 0.0002710089145638394, "tokens_per_sec": 150267.89201518343, "vram_gb": 27.710740089416504, "grad_norm": 0.1767578125, "tokens_processed": 2192834560, "tokens_processed_billions": 2.19283456}
+{"timestamp": 1787541231.9513779, "elapsed_hours": 4.057795241806242, "session_id": 2, "event": "train_step", "step": 33470, "epoch": 1, "loss": 2.2083718943595887, "loss_raw": 2.4288599491119385, "perplexity": 9.10088711885258, "muon_lr": 0.009033157277065076, "adamw_lr": 0.0002709947183119522, "tokens_per_sec": 150271.71902981342, "vram_gb": 27.710740089416504, "grad_norm": 0.158203125, "tokens_processed": 2193489920, "tokens_processed_billions": 2.19348992}
+{"timestamp": 1787541236.0811777, "elapsed_hours": 4.058942408429251, "session_id": 2, "event": "train_step", "step": 33480, "epoch": 1, "loss": 2.1073186016082763, "loss_raw": 2.4467613697052, "perplexity": 8.226154089230342, "muon_lr": 0.009032210554624907, "adamw_lr": 0.0002709663166387471, "tokens_per_sec": 150274.1000369323, "vram_gb": 27.299458026885986, "grad_norm": 0.1435546875, "tokens_processed": 2194145280, "tokens_processed_billions": 2.19414528}
+{"timestamp": 1787541240.0512364, "elapsed_hours": 4.060045202506913, "session_id": 2, "event": "train_step", "step": 33490, "epoch": 1, "loss": 2.010696107149124, "loss_raw": 2.569905996322632, "perplexity": 7.468514424532286, "muon_lr": 0.009031737040608279, "adamw_lr": 0.00027095211121824824, "tokens_per_sec": 150278.1236560338, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 2194800640, "tokens_processed_billions": 2.19480064}
+{"timestamp": 1787541244.0985897, "elapsed_hours": 4.061169467303488, "session_id": 2, "event": "train_step", "step": 33500, "epoch": 1, "loss": 2.0122306108474732, "loss_raw": 2.278702974319458, "perplexity": 7.479983685098726, "muon_lr": 0.00903126342474549, "adamw_lr": 0.0002709379027423646, "tokens_per_sec": 150281.3498870874, "vram_gb": 27.710740089416504, "grad_norm": 0.181640625, "tokens_processed": 2195456000, "tokens_processed_billions": 2.195456}
+{"timestamp": 1787541244.103416, "elapsed_hours": 4.061170808209313, "session_id": 2, "event": "lambda_stats", "step": 33500, "epoch": 1, "lambda_min": 1.109375, "lambda_max": 2.015625, "lambda_mean": 1.1966145833333333, "lambda_std": 0.20269468977080954, "lambda_all": [2.015625, 1.265625, 1.2109375, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.125, 1.140625, 1.109375, 1.140625, 1.109375, 1.125, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787541257.6578157, "elapsed_hours": 4.064935919377539, "session_id": 2, "event": "validation", "step": 33500, "epoch": 1, "val_loss": 2.0166217213869095, "val_perplexity": 7.512901339930947, "is_best": true, "best_loss": 2.0166217213869095, "best_perplexity": 7.512901339930947, "train_loss": 2.278702974319458, "train_loss_avg": 2.0122306108474732}